paper/FinalNtb.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import sys\n",
    "import os\n",
    "sys.path.append('..')\n",
    "import warnings\n",
    "if not sys.warnoptions:\n",
    "    warnings.simplefilter(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from joblib import delayed, Parallel, parallel_backend\n",
    "import time\n",
    "from operator import itemgetter\n",
    "ig0 = itemgetter(0)\n",
    "ig1 = itemgetter(1)\n",
    "ig2 = itemgetter(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import python packages\n",
    "from geomle import DataGenerator, geomle, mle\n",
    "\n",
    "#Import R packages\n",
    "import rpy2\n",
    "import rpy2.robjects as ro\n",
    "from rpy2.robjects.numpy2ri import ri2py, py2ri\n",
    "from rpy2.robjects.packages import importr\n",
    "intdimr = importr('intrinsicDimension')\n",
    "r_base = importr('base')\n",
    "\n",
    "#Import C++ packages\n",
    "import hein"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from tqdm import tqdm\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_time(func): \n",
    "    def inner_func(*args, **kwargs): \n",
    "        begin = time.time() \n",
    "        res = func(*args, **kwargs) \n",
    "        end = time.time()\n",
    "        return res, end - begin\n",
    "    return inner_func"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "# k1_sigmoid = lambda x: int(round(20 / (1 + np.exp( -0.05 * x))))\n",
    "# k2_sigmoid = lambda x: int(round(60 / (1 + np.exp( -0.03 * (x + 5)))))\n",
    "\n",
    "CPU_COUNT = os.cpu_count()\n",
    "k1_log = lambda x: int(np.round(3 * np.log(1/3 * x) + 10))\n",
    "k2_log = lambda x: int(np.round(12 * np.log(1/12 * (x + 10)) + 30))\n",
    "\n",
    "class DimEst():\n",
    "    def __init__(self):\n",
    "        self.names = ['MLE', 'GeoMLE', 'MIND', 'DANCo', 'ESS', 'PCA', 'CD', 'Hein']\n",
    "#         self.names = ['MLE', 'GeoMLE', 'MIND', 'ESS', 'PCA', 'CD', 'Hein']\n",
    "    \n",
    "    def estimateAllMethod(self, data_py):\n",
    "        data_r = py2ri(data_py.values)\n",
    "        dim = data_py.shape[1]\n",
    "        self.funcs = {'MLE':          self.mle(data_py),\n",
    "                      'GeoMLE':       self.geomle(data_py, dim),\n",
    "                      'MiND':         self.mind_mlk(data_r, dim),\n",
    "                      'DANCo':        self.danco(data_r, dim),\n",
    "                      'ESS':          self.ess(data_r),\n",
    "                      'PCA':          self.pca(data_r),\n",
    "                      'CD':           self.cd(data_py),\n",
    "                      'Hein':         self.hein(data_py)}\n",
    "        self.times = {key: ig1(val) for key, val in self.funcs.items()}\n",
    "        self.funcs = {key: ig0(val) for key, val in self.funcs.items()}\n",
    "        return self.funcs, self.times\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def mle(data):\n",
    "        return mle(data, k1=20, k2=55, average=True)[0].mean()\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def geomle(data, dim):\n",
    "#         k1 =  k1_log(dim)\n",
    "#         k2 =  k2_log(dim)\n",
    "        return geomle(data, k1=20, k2=55, nb_iter1=1, alpha=5e-3).mean()\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def mind_mlk(data, dim):\n",
    "        return intdimr.dancoDimEst(data, k=(20 + 55) // 2, D=dim, ver=\"MIND_MLk\").rx2('dim.est')[0]\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def danco(data, dim):\n",
    "        return intdimr.dancoDimEst(data, k=(20 + 55) // 2, D=dim, ver=\"DANCo\").rx2('dim.est')[0]\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def ess(data):\n",
    "        return intdimr.essLocalDimEst(data).rx2('dim.est')[0]\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def pca(data):\n",
    "        return intdimr.pcaLocalDimEst(data, 'FO').rx2('dim.est')[0]\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def hein(data):\n",
    "        res = hein.heindim(data)\n",
    "        return res[0]\n",
    "    \n",
    "    @staticmethod\n",
    "    @calculate_time\n",
    "    def cd(data):\n",
    "        res = hein.heindim(data)\n",
    "        return res[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "DG = DataGenerator()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['Helix1d', 'Helix2d', 'Helicoid', 'Spiral', 'Roll', 'Scurve', 'Star', 'Moebius', 'Sphere', 'Norm', 'Uniform', 'Cubic', 'Affine_3to5', 'Affine', 'Nonlinear_4to6', 'Nonlinear', 'Paraboloid'])"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DG.dict_gen.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'MLE': 4.159317606740128,\n",
       " 'GeoMLE': 5.857447900193141,\n",
       " 'MiND': 4,\n",
       " 'DANCo': 5,\n",
       " 'ESS': 1.4489290575906717,\n",
       " 'PCA': 1,\n",
       " 'CD': 2.708155393600464,\n",
       " 'Hein': 3.0}"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = DG.gen_data('Paraboloid', 1000, 21, 6)\n",
    "DE.estimateAllMethod(data)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'MLE': 3.713450300138347,\n",
       " 'GeoMLE': 4.144922445552533,\n",
       " 'MiND': 4,\n",
       " 'DANCo': 5,\n",
       " 'ESS': 2.998079005110002,\n",
       " 'PCA': 5,\n",
       " 'CD': 3.8828516006469727,\n",
       " 'Hein': 4.0}"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = DG.gen_data('Nonlinear_4to6', 1000, 6, 4)\n",
    "DE.estimateAllMethod(data)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DE.estimateAllMethod(data)[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'MLE': 3.8538676445163946,\n",
       " 'GeoMLE': 3.9457740204973577,\n",
       " 'MiND': 4,\n",
       " 'DANCo': 4,\n",
       " 'ESS': 5.117706889110422,\n",
       " 'PCA': 5,\n",
       " 'CD': 3.7512459754943848,\n",
       " 'Hein': 4.0}"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = DG.gen_data('Sphere', 1000, 6, 4, 'uniform')\n",
    "DE.estimateAllMethod(data)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'MLE': 3.6840320886740607,\n",
       " 'GeoMLE': 3.9836083647587768,\n",
       " 'MiND': 4,\n",
       " 'DANCo': 5,\n",
       " 'ESS': 4.968565931670556,\n",
       " 'PCA': 5,\n",
       " 'CD': 3.651015520095825,\n",
       " 'Hein': 4.0}"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = DG.gen_data('Sphere', 1000, 6, 4, 'nonuniform')\n",
    "DE.estimateAllMethod(data)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'MLE': 22.760918003287422,\n",
       " 'GeoMLE': 42.04015022578717,\n",
       " 'MiND': 25,\n",
       " 'DANCo': 41,\n",
       " 'ESS': 41.00432259127065,\n",
       " 'PCA': 41,\n",
       " 'CD': 26.72852325439453,\n",
       " 'Hein': 27.0}"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = DG.gen_data('Sphere', 1000, 60, 40, 'nonuniform')\n",
    "DE.estimateAllMethod(data)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decribing algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explain k1 and k2 dependancy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "ds = np.arange(2, 150)\n",
    "k1_log_arr = np.array(list(map(k1_log, ds)))\n",
    "k2_log_arr = np.array(list(map(k2_log, ds)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f55152ef550>]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(ds, k1_log_arr)\n",
    "plt.plot(ds, k2_log_arr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explain main idea of algorithms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Ridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = []\n",
    "name = 'Sphere'\n",
    "N = 1000\n",
    "ds = [6, 7, 8, 9]\n",
    "\n",
    "for d in ds:\n",
    "    data = DG.gen_data(name, N, d+1, d)\n",
    "    res, id_data = geomle(data, k1=7, k2=40, nb_iter1=1, debug=True)\n",
    "    df.append(id_data[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(1, 1, figsize=(10, 8))\n",
    "\n",
    "\n",
    "alpha_poly = (1, 2)\n",
    "alpha_reg = 1e-2\n",
    "\n",
    "cm = plt.cm.get_cmap('Set1', 10)\n",
    "save_data = []\n",
    "\n",
    "\n",
    "for i in range(len(df)):\n",
    "    gr_df = df[i].groupby(['k', 'idx']).mean().reset_index().groupby('k')\n",
    "\n",
    "    x = gr_df['R'].mean().values\n",
    "    y = gr_df['dim'].mean().values\n",
    "    w = gr_df['dim'].std().values\n",
    "    save_data.append([x, y, w])\n",
    "    X = np.array([x ** i for i in alpha_poly]).T\n",
    "\n",
    "    xlin = np.linspace(0, x.max(), 100)\n",
    "    Xlin = np.array([xlin ** i for i in alpha_poly]).T\n",
    "    lm = Ridge(alpha=alpha_reg)\n",
    "    lm.fit(X, y, w **-1)\n",
    "    plt.errorbar(x, y, w * 0.1, fmt='o', color=cm(i), capsize=3, capthick=0.5)\n",
    "    plt.plot(xlin, lm.predict(Xlin), color=cm(i))\n",
    "\n",
    "# ax.legend()\n",
    "ax.grid()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('result/exp_new.npy', save_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Density test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gen_sphere_data_with_distribution(n, dim, d, distrib):\n",
    "    if distrib == 'uniform':\n",
    "        V = np.random.randn(n,d + 1)\n",
    "    elif distrib == 'nonuniform':\n",
    "        V = (np.random.rand(n,d + 1) - 0.5) * 3\n",
    "    data = pd.DataFrame(np.hstack([V/np.sqrt(np.sum(V**2, axis=1))[:,None],\n",
    "                                          np.zeros((n, dim - d - 1))]))\n",
    "    assert data.shape == (n, dim)\n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 5))\n",
    "plt.subplot(121)\n",
    "data = gen_sphere_data_with_distribution(5000, 5, 4, 'uniform')\n",
    "plt.hist2d(data[0], data[1], bins=30)\n",
    "\n",
    "plt.subplot(122)\n",
    "data = gen_sphere_data_with_distribution(5000, 5, 4, 'nonuniform')\n",
    "plt.hist2d(data[0], data[1], bins=30)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 10/10 [00:20<00:00,  2.09s/it]\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Uniform\n",
      "MLE:    4.8659(0.0708)\n",
      "GeoMLE: 5.1343(0.1778)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 10/10 [00:20<00:00,  2.08s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NonUniform\n",
      "MLE:    4.5906(0.0485)\n",
      "GeoMLE: 4.9855(0.1169)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "res = []\n",
    "for i in tqdm(range(10)):\n",
    "    data = gen_sphere_data_with_distribution(1000, 7, 5, 'uniform')\n",
    "    res.append([mle(data)[0].mean(), geomle(data, nb_iter1=1, k1=10, k2=40).mean()])\n",
    "                \n",
    "mean = np.mean(res, axis=0)\n",
    "std = np.std(res, axis=0)\n",
    "print('Uniform')\n",
    "print('MLE:    {:.4f}({:.4f})'.format(mean[0], std[0]))\n",
    "print('GeoMLE: {:.4f}({:.4f})'.format(mean[1], std[1]))\n",
    "\n",
    "\n",
    "res = []\n",
    "for i in tqdm(range(10)):\n",
    "    data = gen_sphere_data_with_distribution(1000, 7, 5, 'nonuniform')\n",
    "    res.append([mle(data)[0].mean(), geomle(data, nb_iter1=1, k1=10, k2=40).mean()])\n",
    "                \n",
    "mean = np.mean(res, axis=0)\n",
    "std = np.std(res, axis=0)\n",
    "print('NonUniform')\n",
    "print('MLE:    {:.4f}({:.4f})'.format(mean[0], std[0]))\n",
    "print('GeoMLE: {:.4f}({:.4f})'.format(mean[1], std[1]))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Funcs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_one_test(args):\n",
    "    try:\n",
    "        data = DG.gen_data(name = args['name'],\n",
    "                           n = args['N'],\n",
    "                           dim = args['D'],\n",
    "                           d = args['d'],\n",
    "                           type_sample = args['type_sample'],\n",
    "                           noise = args['noise'])\n",
    "        res, t = DE.estimateAllMethod(data)\n",
    "        res['Dataset'] = args['name']\n",
    "        res['Num'] = args['N']\n",
    "        res['Dim'] = args['D']\n",
    "        res['RealDim'] = args['d']\n",
    "        t['Dataset'] = args['name']\n",
    "        t['Num'] = args['N']\n",
    "        t['Dim'] = args['D']\n",
    "        t['RealDim'] = args['d']\n",
    "        return list(res.values()), list(t.values())\n",
    "    except:\n",
    "        print('WRONG with {}'.format(args))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_test(func, steps, const_args, n_repetition, n_jobs=CPU_COUNT, return_time=False):\n",
    "    with parallel_backend('multiprocessing'):\n",
    "        result = Parallel(n_jobs=n_jobs)(delayed(func)(add_args(const_args, step))\n",
    "                                 for step in tqdm(steps * n_repetition))\n",
    "    result = [i for i in result if i is not None]\n",
    "    times_res = np.array(list(map(ig1, result)))\n",
    "    dims_res = np.array(list(map(ig0, result)))\n",
    "    cols = DE.names + ['Dataset', 'Num', 'Dim', 'RealDim']\n",
    "    df_res = df = pd.DataFrame(dims_res, columns=cols)\n",
    "    df_time = pd.DataFrame(times_res, columns=cols)\n",
    "    df_res = df_res.set_index('Dataset').astype('float').reset_index()\n",
    "    df_time = df_time.set_index('Dataset').astype('float').reset_index()\n",
    "    return df_res, df_time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dependacne of manifold dimension"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()\n",
    "DG = DataGenerator()\n",
    "DG.dict_gen.keys()\n",
    "\n",
    "steps = np.arange(2, 100, 4).tolist()\n",
    "args = {'name': 'Sphere',\n",
    "        'D': 100,\n",
    "        'N': 1000,\n",
    "        'type_sample': 'uniform',\n",
    "        'noise': 0}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_args(const_args, step):\n",
    "    args = {**const_args}\n",
    "    args['d'] = step\n",
    "    assert len({'name', 'D', 'd', 'N', 'type_sample', 'noise'} & set(args.keys())) == 6\n",
    "    return args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 125/125 [08:38<00:00,  4.34s/it]\n"
     ]
    }
   ],
   "source": [
    "names = DE.names\n",
    "n_repetition = 5\n",
    "\n",
    "df, _ = run_test(run_one_test, steps, args, n_repetition)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_mean = df.groupby(['Dataset', 'Num', 'Dim', 'RealDim']).mean().reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(1, 1, figsize=(10, 8))\n",
    "\n",
    "ax.plot(df_mean['RealDim'], df_mean['RealDim'], '--', color='black')\n",
    "for name in names:\n",
    "    ax.plot(df_mean['RealDim'], df_mean[name], label=name)\n",
    "    \n",
    "ax.legend()\n",
    "ax.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('result/result_sphere.csv', index_label=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dolan More Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()\n",
    "DG = DataGenerator()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['Helix1d', 'Helix2d', 'Helicoid', 'Spiral', 'Roll', 'Scurve', 'Star', 'Moebius', 'Sphere', 'Norm', 'Uniform', 'Cubic', 'Affine_3to5', 'Affine', 'Nonlinear_4to6', 'Nonlinear', 'Paraboloid'])"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DG.dict_gen.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "steps = [\n",
    "    ('Sphere',  5,  3),\n",
    "    ('Sphere',  10, 5),\n",
    "    ('Sphere',  15, 10),\n",
    "    ('Sphere',  35, 30),\n",
    "    ('Sphere',  55, 50),\n",
    "    ('Sphere',  75, 70),\n",
    "    \n",
    "    ('Affine_3to5',  5,  3),\n",
    "    ('Affine',       5,  5),\n",
    "    ('Affine',       10, 10),\n",
    "    ('Affine',       20, 20),\n",
    "    \n",
    "    ('Nonlinear_4to6',  6,  4),\n",
    "    ('Nonlinear',       8,  4),\n",
    "    ('Nonlinear',       36, 6),\n",
    "    ('Nonlinear',       64, 8),\n",
    "    ('Nonlinear',       72, 12),\n",
    "    \n",
    "    ('Spiral',   3,  1),\n",
    "    ('Spiral',   13, 1),\n",
    "    ('Helix1d',  3,  1),\n",
    "    ('Helix1d',  13, 1),\n",
    "    ('Helix2d',  3,  2),\n",
    "    ('Helix2d',  13, 2),\n",
    "    ('Helicoid', 3,  2),\n",
    "    ('Helicoid', 13, 2),\n",
    "    ('Roll',     3,  2),\n",
    "    ('Roll',     13, 2),\n",
    "    ('Scurve',   3,  2),\n",
    "    ('Scurve',   13, 2),\n",
    "    ('Moebius',  3,  2),\n",
    "    \n",
    "    ('Uniform', 10, 5),\n",
    "    ('Uniform', 15, 10),\n",
    "    ('Uniform', 35, 30),\n",
    "    ('Uniform', 55, 50),\n",
    "    ('Uniform', 75, 70),\n",
    "     \n",
    "    ('Norm', 20, 20),\n",
    "    ('Norm', 50, 20),\n",
    "    ('Norm', 50, 50),\n",
    "    ('Norm', 70, 50),\n",
    "    \n",
    "    ('Paraboloid', 12, 3),\n",
    "    ('Paraboloid', 21, 6),\n",
    "    ('Paraboloid', 30, 9),\n",
    "    ('Paraboloid', 39, 12),\n",
    "    ('Paraboloid', 48, 15),\n",
    "    \n",
    "    ('Cubic', 5,  3),\n",
    "    ('Cubic', 35, 30),\n",
    "    ('Cubic', 90, 70),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_args(const_args, step):\n",
    "    args = {**const_args}\n",
    "    args['name'] = step[0]\n",
    "    args['D'] = step[1]\n",
    "    args['d'] = step[2]\n",
    "    assert len({'name', 'D', 'd', 'N', 'noise'} & set(args.keys())) == 5\n",
    "    return args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 96%|█████████▋| 217/225 [19:31<00:37,  4.68s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\n",
      " 1\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 97%|█████████▋| 218/225 [19:36<00:32,  4.66s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'N': 1000, 'noise': 0, 'type_sample': 'uniform', 'name': 'Roll', 'D': 13, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 225/225 [20:07<00:00,  6.55s/it]\n"
     ]
    }
   ],
   "source": [
    "n_repetition = 5\n",
    "args = {'N': 1000, 'noise': 0, 'type_sample': 'uniform'}\n",
    "df, df_time = run_test(run_one_test, steps, args, n_repetition)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>MLE</th>\n",
       "      <th>GeoMLE</th>\n",
       "      <th>MIND</th>\n",
       "      <th>DANCo</th>\n",
       "      <th>ESS</th>\n",
       "      <th>PCA</th>\n",
       "      <th>CD</th>\n",
       "      <th>Hein</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dataset</th>\n",
       "      <th>Num</th>\n",
       "      <th>Dim</th>\n",
       "      <th>RealDim</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Affine</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.305258</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.978156</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.574749</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>7.712303</td>\n",
       "      <td>9.912339</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.6</td>\n",
       "      <td>9.967358</td>\n",
       "      <td>10.0</td>\n",
       "      <td>8.314556</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20.0</th>\n",
       "      <th>20.0</th>\n",
       "      <td>13.310821</td>\n",
       "      <td>19.687696</td>\n",
       "      <td>14.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>20.069427</td>\n",
       "      <td>20.0</td>\n",
       "      <td>14.577997</td>\n",
       "      <td>14.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Affine_3to5</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.738002</td>\n",
       "      <td>3.072325</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.592155</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.888834</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Cubic</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>3.046447</td>\n",
       "      <td>3.117651</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.990806</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.036158</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>18.235634</td>\n",
       "      <td>29.721372</td>\n",
       "      <td>19.2</td>\n",
       "      <td>29.8</td>\n",
       "      <td>30.631984</td>\n",
       "      <td>31.0</td>\n",
       "      <td>20.529007</td>\n",
       "      <td>21.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>31.211757</td>\n",
       "      <td>74.293434</td>\n",
       "      <td>33.8</td>\n",
       "      <td>69.6</td>\n",
       "      <td>71.025555</td>\n",
       "      <td>71.0</td>\n",
       "      <td>35.243558</td>\n",
       "      <td>37.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helicoid</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.139040</td>\n",
       "      <td>2.314091</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.834995</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.938105</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.118204</td>\n",
       "      <td>2.222067</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.839164</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.022425</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helix1d</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.088490</td>\n",
       "      <td>1.216671</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.632561</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.002036</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.086904</td>\n",
       "      <td>1.278032</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.636503</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.989992</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helix2d</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.764936</td>\n",
       "      <td>2.677941</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.564205</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.973622</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.767395</td>\n",
       "      <td>2.810647</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.541431</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.041041</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moebius</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.091290</td>\n",
       "      <td>1.278846</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.322124</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.040221</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">Nonlinear</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">1000.0</th>\n",
       "      <th>8.0</th>\n",
       "      <th>4.0</th>\n",
       "      <td>4.340946</td>\n",
       "      <td>3.320108</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.865763</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.815609</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36.0</th>\n",
       "      <th>6.0</th>\n",
       "      <td>6.943705</td>\n",
       "      <td>6.708093</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.8</td>\n",
       "      <td>12.080227</td>\n",
       "      <td>12.0</td>\n",
       "      <td>5.776229</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64.0</th>\n",
       "      <th>8.0</th>\n",
       "      <td>9.252331</td>\n",
       "      <td>10.690596</td>\n",
       "      <td>9.0</td>\n",
       "      <td>11.2</td>\n",
       "      <td>15.923075</td>\n",
       "      <td>16.0</td>\n",
       "      <td>7.912540</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72.0</th>\n",
       "      <th>12.0</th>\n",
       "      <td>12.974117</td>\n",
       "      <td>17.306892</td>\n",
       "      <td>13.2</td>\n",
       "      <td>17.8</td>\n",
       "      <td>24.007460</td>\n",
       "      <td>24.0</td>\n",
       "      <td>12.154475</td>\n",
       "      <td>12.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Nonlinear_4to6</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>6.0</th>\n",
       "      <th>4.0</th>\n",
       "      <td>3.700495</td>\n",
       "      <td>4.127521</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.062024</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.609223</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">Norm</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">1000.0</th>\n",
       "      <th>20.0</th>\n",
       "      <th>20.0</th>\n",
       "      <td>13.873363</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>19.785223</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.699330</td>\n",
       "      <td>14.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">50.0</th>\n",
       "      <th>20.0</th>\n",
       "      <td>13.767693</td>\n",
       "      <td>21.519287</td>\n",
       "      <td>15.0</td>\n",
       "      <td>20.6</td>\n",
       "      <td>20.005668</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.841544</td>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50.0</th>\n",
       "      <td>24.476454</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>27.0</td>\n",
       "      <td>47.8</td>\n",
       "      <td>50.016444</td>\n",
       "      <td>50.0</td>\n",
       "      <td>24.326109</td>\n",
       "      <td>26.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>24.441226</td>\n",
       "      <td>52.875505</td>\n",
       "      <td>27.0</td>\n",
       "      <td>47.4</td>\n",
       "      <td>49.685605</td>\n",
       "      <td>50.0</td>\n",
       "      <td>24.307200</td>\n",
       "      <td>26.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">Paraboloid</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1000.0</th>\n",
       "      <th>12.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.734067</td>\n",
       "      <td>3.062384</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.804378</td>\n",
       "      <td>1.2</td>\n",
       "      <td>2.041507</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21.0</th>\n",
       "      <th>6.0</th>\n",
       "      <td>4.202450</td>\n",
       "      <td>5.961053</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.471528</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.677824</td>\n",
       "      <td>2.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30.0</th>\n",
       "      <th>9.0</th>\n",
       "      <td>4.801959</td>\n",
       "      <td>9.033189</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.304623</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.898638</td>\n",
       "      <td>2.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39.0</th>\n",
       "      <th>12.0</th>\n",
       "      <td>4.934400</td>\n",
       "      <td>12.250517</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1.222877</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.263406</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48.0</th>\n",
       "      <th>15.0</th>\n",
       "      <td>4.721062</td>\n",
       "      <td>13.886602</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.8</td>\n",
       "      <td>1.166538</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.264442</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Roll</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.956159</td>\n",
       "      <td>2.044432</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.958000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.983678</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.952086</td>\n",
       "      <td>2.207737</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.977422</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.993772</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Scurve</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.912767</td>\n",
       "      <td>2.041083</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.220897</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.985736</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.912798</td>\n",
       "      <td>2.011035</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.179040</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.926215</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">Sphere</th>\n",
       "      <th rowspan=\"6\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.972343</td>\n",
       "      <td>3.081803</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.967764</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.006982</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.759788</td>\n",
       "      <td>5.092993</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.027623</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.028022</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>8.533141</td>\n",
       "      <td>10.377820</td>\n",
       "      <td>9.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>11.071274</td>\n",
       "      <td>11.0</td>\n",
       "      <td>9.371253</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>18.985804</td>\n",
       "      <td>32.463958</td>\n",
       "      <td>20.4</td>\n",
       "      <td>30.8</td>\n",
       "      <td>30.969825</td>\n",
       "      <td>31.0</td>\n",
       "      <td>23.095783</td>\n",
       "      <td>23.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>26.528460</td>\n",
       "      <td>52.898723</td>\n",
       "      <td>28.4</td>\n",
       "      <td>51.2</td>\n",
       "      <td>51.089642</td>\n",
       "      <td>51.0</td>\n",
       "      <td>31.113284</td>\n",
       "      <td>32.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>32.671610</td>\n",
       "      <td>70.743789</td>\n",
       "      <td>35.2</td>\n",
       "      <td>70.4</td>\n",
       "      <td>70.533346</td>\n",
       "      <td>71.0</td>\n",
       "      <td>38.923859</td>\n",
       "      <td>40.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Spiral</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.740074</td>\n",
       "      <td>1.547418</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.084984</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.972060</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.736327</td>\n",
       "      <td>1.559839</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.098838</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.992213</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">Uniform</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1000.0</th>\n",
       "      <th>10.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.310223</td>\n",
       "      <td>4.740012</td>\n",
       "      <td>4.4</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.010652</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.776037</td>\n",
       "      <td>4.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>7.787836</td>\n",
       "      <td>9.176881</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.8</td>\n",
       "      <td>10.115493</td>\n",
       "      <td>10.0</td>\n",
       "      <td>8.425551</td>\n",
       "      <td>8.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>17.839916</td>\n",
       "      <td>29.365858</td>\n",
       "      <td>19.0</td>\n",
       "      <td>28.8</td>\n",
       "      <td>29.934330</td>\n",
       "      <td>30.0</td>\n",
       "      <td>19.509309</td>\n",
       "      <td>20.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>24.935675</td>\n",
       "      <td>51.173918</td>\n",
       "      <td>27.0</td>\n",
       "      <td>49.0</td>\n",
       "      <td>49.359511</td>\n",
       "      <td>50.0</td>\n",
       "      <td>28.217013</td>\n",
       "      <td>30.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>31.016346</td>\n",
       "      <td>73.698438</td>\n",
       "      <td>33.8</td>\n",
       "      <td>69.2</td>\n",
       "      <td>70.855069</td>\n",
       "      <td>70.0</td>\n",
       "      <td>35.588058</td>\n",
       "      <td>38.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          MLE     GeoMLE  MIND  DANCo  \\\n",
       "Dataset        Num    Dim  RealDim                                      \n",
       "Affine         1000.0 5.0  5.0       4.305258   5.000000   4.0    5.0   \n",
       "                      10.0 10.0      7.712303   9.912339   8.0    9.6   \n",
       "                      20.0 20.0     13.310821  19.687696  14.0   19.0   \n",
       "Affine_3to5    1000.0 5.0  3.0       2.738002   3.072325   3.0    3.0   \n",
       "Cubic          1000.0 5.0  3.0       3.046447   3.117651   3.0    4.0   \n",
       "                      35.0 30.0     18.235634  29.721372  19.2   29.8   \n",
       "                      90.0 70.0     31.211757  74.293434  33.8   69.6   \n",
       "Helicoid       1000.0 3.0  2.0       2.139040   2.314091   2.0    2.0   \n",
       "                      13.0 2.0       2.118204   2.222067   2.0    2.0   \n",
       "Helix1d        1000.0 3.0  1.0       1.088490   1.216671   1.0    1.0   \n",
       "                      13.0 1.0       1.086904   1.278032   1.0    1.0   \n",
       "Helix2d        1000.0 3.0  2.0       2.764936   2.677941   3.0    3.0   \n",
       "                      13.0 2.0       2.767395   2.810647   3.0    3.0   \n",
       "Moebius        1000.0 3.0  2.0       2.091290   1.278846   2.0    2.0   \n",
       "Nonlinear      1000.0 8.0  4.0       4.340946   3.320108   4.0    5.0   \n",
       "                      36.0 6.0       6.943705   6.708093   7.0    7.8   \n",
       "                      64.0 8.0       9.252331  10.690596   9.0   11.2   \n",
       "                      72.0 12.0     12.974117  17.306892  13.2   17.8   \n",
       "Nonlinear_4to6 1000.0 6.0  4.0       3.700495   4.127521   4.0    4.0   \n",
       "Norm           1000.0 20.0 20.0     13.873363  20.000000  15.0   20.0   \n",
       "                      50.0 20.0     13.767693  21.519287  15.0   20.6   \n",
       "                           50.0     24.476454  50.000000  27.0   47.8   \n",
       "                      70.0 50.0     24.441226  52.875505  27.0   47.4   \n",
       "Paraboloid     1000.0 12.0 3.0       2.734067   3.062384   3.0    3.2   \n",
       "                      21.0 6.0       4.202450   5.961053   5.0    5.0   \n",
       "                      30.0 9.0       4.801959   9.033189   5.0    6.0   \n",
       "                      39.0 12.0      4.934400  12.250517   6.0    7.0   \n",
       "                      48.0 15.0      4.721062  13.886602   6.0    6.8   \n",
       "Roll           1000.0 3.0  2.0       1.956159   2.044432   2.0    2.0   \n",
       "                      13.0 2.0       1.952086   2.207737   2.0    2.0   \n",
       "Scurve         1000.0 3.0  2.0       1.912767   2.041083   2.0    2.0   \n",
       "                      13.0 2.0       1.912798   2.011035   2.0    2.0   \n",
       "Sphere         1000.0 5.0  3.0       2.972343   3.081803   3.0    4.0   \n",
       "                      10.0 5.0       4.759788   5.092993   5.0    6.0   \n",
       "                      15.0 10.0      8.533141  10.377820   9.0   11.0   \n",
       "                      35.0 30.0     18.985804  32.463958  20.4   30.8   \n",
       "                      55.0 50.0     26.528460  52.898723  28.4   51.2   \n",
       "                      75.0 70.0     32.671610  70.743789  35.2   70.4   \n",
       "Spiral         1000.0 3.0  1.0       1.740074   1.547418   2.0    2.0   \n",
       "                      13.0 1.0       1.736327   1.559839   2.0    2.0   \n",
       "Uniform        1000.0 10.0 5.0       4.310223   4.740012   4.4    5.0   \n",
       "                      15.0 10.0      7.787836   9.176881   8.0    9.8   \n",
       "                      35.0 30.0     17.839916  29.365858  19.0   28.8   \n",
       "                      55.0 50.0     24.935675  51.173918  27.0   49.0   \n",
       "                      75.0 70.0     31.016346  73.698438  33.8   69.2   \n",
       "\n",
       "                                          ESS   PCA         CD  Hein  \n",
       "Dataset        Num    Dim  RealDim                                    \n",
       "Affine         1000.0 5.0  5.0       4.978156   5.0   4.574749   5.0  \n",
       "                      10.0 10.0      9.967358  10.0   8.314556   8.0  \n",
       "                      20.0 20.0     20.069427  20.0  14.577997  14.8  \n",
       "Affine_3to5    1000.0 5.0  3.0       2.592155   3.0   2.888834   3.0  \n",
       "Cubic          1000.0 5.0  3.0       3.990806   4.0   3.036158   3.0  \n",
       "                      35.0 30.0     30.631984  31.0  20.529007  21.0  \n",
       "                      90.0 70.0     71.025555  71.0  35.243558  37.2  \n",
       "Helicoid       1000.0 3.0  2.0       1.834995   2.0   1.938105   2.0  \n",
       "                      13.0 2.0       1.839164   2.0   2.022425   2.0  \n",
       "Helix1d        1000.0 3.0  1.0       2.632561   3.0   1.002036   1.0  \n",
       "                      13.0 1.0       2.636503   3.0   0.989992   1.0  \n",
       "Helix2d        1000.0 3.0  2.0       2.564205   3.0   1.973622   2.0  \n",
       "                      13.0 2.0       2.541431   3.0   2.041041   2.0  \n",
       "Moebius        1000.0 3.0  2.0       2.322124   3.0   2.040221   2.0  \n",
       "Nonlinear      1000.0 8.0  4.0       7.865763   8.0   3.815609   4.0  \n",
       "                      36.0 6.0      12.080227  12.0   5.776229   6.0  \n",
       "                      64.0 8.0      15.923075  16.0   7.912540   8.0  \n",
       "                      72.0 12.0     24.007460  24.0  12.154475  12.2  \n",
       "Nonlinear_4to6 1000.0 6.0  4.0       3.062024   5.0   3.609223   4.0  \n",
       "Norm           1000.0 20.0 20.0     19.785223  20.0  13.699330  14.6  \n",
       "                      50.0 20.0     20.005668  20.0  13.841544  14.0  \n",
       "                           50.0     50.016444  50.0  24.326109  26.2  \n",
       "                      70.0 50.0     49.685605  50.0  24.307200  26.0  \n",
       "Paraboloid     1000.0 12.0 3.0       1.804378   1.2   2.041507   2.0  \n",
       "                      21.0 6.0       1.471528   1.0   2.677824   2.2  \n",
       "                      30.0 9.0       1.304623   1.0   2.898638   2.6  \n",
       "                      39.0 12.0      1.222877   1.0   3.263406   3.0  \n",
       "                      48.0 15.0      1.166538   1.0   3.264442   3.0  \n",
       "Roll           1000.0 3.0  2.0       2.958000   3.0   1.983678   2.0  \n",
       "                      13.0 2.0       2.977422   3.0   1.993772   2.0  \n",
       "Scurve         1000.0 3.0  2.0       2.220897   3.0   1.985736   2.0  \n",
       "                      13.0 2.0       2.179040   3.0   1.926215   2.0  \n",
       "Sphere         1000.0 5.0  3.0       3.967764   4.0   3.006982   3.0  \n",
       "                      10.0 5.0       6.027623   6.0   5.028022   5.0  \n",
       "                      15.0 10.0     11.071274  11.0   9.371253   9.0  \n",
       "                      35.0 30.0     30.969825  31.0  23.095783  23.4  \n",
       "                      55.0 50.0     51.089642  51.0  31.113284  32.4  \n",
       "                      75.0 70.0     70.533346  71.0  38.923859  40.4  \n",
       "Spiral         1000.0 3.0  1.0       2.084984   2.0   0.972060   1.0  \n",
       "                      13.0 1.0       2.098838   2.0   0.992213   1.0  \n",
       "Uniform        1000.0 10.0 5.0       5.010652   5.0   4.776037   4.8  \n",
       "                      15.0 10.0     10.115493  10.0   8.425551   8.6  \n",
       "                      35.0 30.0     29.934330  30.0  19.509309  20.6  \n",
       "                      55.0 50.0     49.359511  50.0  28.217013  30.4  \n",
       "                      75.0 70.0     70.855069  70.0  35.588058  38.2  "
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['Dataset', 'Num', 'Dim', 'RealDim']).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dfr = df.copy().round()\n",
    "names = dfr.columns.difference(['Dataset', 'Num', 'Dim', 'RealDim'])\n",
    "\n",
    "f, ax = plt.subplots(1, 1, figsize=(12, 8))\n",
    "nsteps = 5001\n",
    "max_tau = 40\n",
    "errors_total = abs(dfr[names].sub(dfr['RealDim'], axis=0).values)\n",
    "errors_total = np.max([errors_total, np.ones_like(errors_total)], axis=0)\n",
    "\n",
    "\n",
    "taus = np.linspace(1, max_tau, nsteps)        \n",
    "rhos = np.zeros((nsteps, len(names)))\n",
    "for cnt, tau in enumerate(taus):\n",
    "    rhos[cnt] = np.sum(errors_total <= tau, axis=0)\n",
    "\n",
    "for i in range(len(names)):\n",
    "    ax.plot(taus, rhos[:,i]/errors_total.shape[0], label=names[i])\n",
    "\n",
    "ax.set_title('Dolan-More Curve')\n",
    "ax.legend()\n",
    "ax.grid()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CD        0.193988\n",
       "DANCo     0.164122\n",
       "ESS       0.384890\n",
       "GeoMLE    0.108177\n",
       "Hein      0.188248\n",
       "MIND      0.248055\n",
       "MLE       0.269471\n",
       "PCA       0.421967\n",
       "dtype: float64"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "abs(dfr[names].sub(dfr['RealDim'], axis=0)).mul(dfr['RealDim']**-1, axis=0).mean(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('result/result_dm_new.csv', index_label=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>MLE</th>\n",
       "      <th>GeoMLE</th>\n",
       "      <th>MIND</th>\n",
       "      <th>DANCo</th>\n",
       "      <th>ESS</th>\n",
       "      <th>PCA</th>\n",
       "      <th>CD</th>\n",
       "      <th>Hein</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dataset</th>\n",
       "      <th>Num</th>\n",
       "      <th>Dim</th>\n",
       "      <th>RealDim</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Affine</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.436176</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.7</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.961017</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.737116</td>\n",
       "      <td>4.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>8.092389</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>8.1</td>\n",
       "      <td>9.5</td>\n",
       "      <td>10.020648</td>\n",
       "      <td>10.0</td>\n",
       "      <td>8.609721</td>\n",
       "      <td>8.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20.0</th>\n",
       "      <th>20.0</th>\n",
       "      <td>14.110037</td>\n",
       "      <td>19.837311</td>\n",
       "      <td>14.5</td>\n",
       "      <td>19.3</td>\n",
       "      <td>20.012921</td>\n",
       "      <td>20.0</td>\n",
       "      <td>14.791872</td>\n",
       "      <td>15.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Affine_3to5</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.819162</td>\n",
       "      <td>3.125630</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.581466</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.917733</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Cubic</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>3.069715</td>\n",
       "      <td>3.120422</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.953560</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.058121</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>19.548624</td>\n",
       "      <td>30.146128</td>\n",
       "      <td>20.5</td>\n",
       "      <td>30.8</td>\n",
       "      <td>30.946966</td>\n",
       "      <td>31.0</td>\n",
       "      <td>20.579042</td>\n",
       "      <td>21.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>34.217167</td>\n",
       "      <td>73.697099</td>\n",
       "      <td>35.9</td>\n",
       "      <td>73.4</td>\n",
       "      <td>71.543753</td>\n",
       "      <td>71.0</td>\n",
       "      <td>34.738716</td>\n",
       "      <td>37.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helicoid</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.086243</td>\n",
       "      <td>2.263129</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.839381</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.031638</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.079526</td>\n",
       "      <td>2.286545</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.833291</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.033630</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helix1d</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.002463</td>\n",
       "      <td>1.399572</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.641996</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.024649</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.002080</td>\n",
       "      <td>1.403783</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.613153</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.006132</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helix2d</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.741861</td>\n",
       "      <td>2.772936</td>\n",
       "      <td>2.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.550043</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.017534</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.743644</td>\n",
       "      <td>2.833216</td>\n",
       "      <td>2.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.547343</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.022974</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moebius</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.985567</td>\n",
       "      <td>1.271459</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.314794</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.020746</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">Nonlinear</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">1000.0</th>\n",
       "      <th>8.0</th>\n",
       "      <th>4.0</th>\n",
       "      <td>4.076561</td>\n",
       "      <td>3.409430</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.7</td>\n",
       "      <td>7.942824</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.760750</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36.0</th>\n",
       "      <th>6.0</th>\n",
       "      <td>6.683937</td>\n",
       "      <td>6.807444</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.1</td>\n",
       "      <td>11.961104</td>\n",
       "      <td>12.0</td>\n",
       "      <td>5.746971</td>\n",
       "      <td>5.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64.0</th>\n",
       "      <th>8.0</th>\n",
       "      <td>9.210539</td>\n",
       "      <td>10.525456</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10.5</td>\n",
       "      <td>16.027877</td>\n",
       "      <td>16.0</td>\n",
       "      <td>7.982724</td>\n",
       "      <td>7.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72.0</th>\n",
       "      <th>12.0</th>\n",
       "      <td>13.437732</td>\n",
       "      <td>17.379927</td>\n",
       "      <td>13.6</td>\n",
       "      <td>17.5</td>\n",
       "      <td>23.923810</td>\n",
       "      <td>24.0</td>\n",
       "      <td>12.080877</td>\n",
       "      <td>11.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Nonlinear_4to6</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>6.0</th>\n",
       "      <th>4.0</th>\n",
       "      <td>3.825456</td>\n",
       "      <td>4.138118</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.144661</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.678472</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">Norm</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">1000.0</th>\n",
       "      <th>20.0</th>\n",
       "      <th>20.0</th>\n",
       "      <td>14.988604</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>15.6</td>\n",
       "      <td>20.0</td>\n",
       "      <td>20.045761</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.565712</td>\n",
       "      <td>14.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">50.0</th>\n",
       "      <th>20.0</th>\n",
       "      <td>14.989692</td>\n",
       "      <td>22.124816</td>\n",
       "      <td>15.8</td>\n",
       "      <td>21.4</td>\n",
       "      <td>20.271046</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.780381</td>\n",
       "      <td>14.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50.0</th>\n",
       "      <td>27.026298</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>28.6</td>\n",
       "      <td>47.5</td>\n",
       "      <td>49.690825</td>\n",
       "      <td>50.0</td>\n",
       "      <td>24.035341</td>\n",
       "      <td>26.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>27.168977</td>\n",
       "      <td>53.282319</td>\n",
       "      <td>28.7</td>\n",
       "      <td>52.9</td>\n",
       "      <td>49.868543</td>\n",
       "      <td>50.0</td>\n",
       "      <td>24.444687</td>\n",
       "      <td>26.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">Paraboloid</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1000.0</th>\n",
       "      <th>12.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.830874</td>\n",
       "      <td>2.999793</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.786439</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.196276</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21.0</th>\n",
       "      <th>6.0</th>\n",
       "      <td>4.707588</td>\n",
       "      <td>6.011846</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.4</td>\n",
       "      <td>1.472586</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.561571</td>\n",
       "      <td>2.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30.0</th>\n",
       "      <th>9.0</th>\n",
       "      <td>5.882575</td>\n",
       "      <td>9.163055</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1.308293</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.766422</td>\n",
       "      <td>2.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39.0</th>\n",
       "      <th>12.0</th>\n",
       "      <td>6.611216</td>\n",
       "      <td>12.152166</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.3</td>\n",
       "      <td>1.223703</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.025258</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48.0</th>\n",
       "      <th>15.0</th>\n",
       "      <td>6.974033</td>\n",
       "      <td>13.636721</td>\n",
       "      <td>7.4</td>\n",
       "      <td>8.7</td>\n",
       "      <td>1.161489</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.327201</td>\n",
       "      <td>3.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Roll</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.954331</td>\n",
       "      <td>2.275757</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.952127</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.998514</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.947018</td>\n",
       "      <td>2.317430</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.974154</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.030448</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Scurve</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.946351</td>\n",
       "      <td>2.037241</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.182087</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.939938</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>1.948070</td>\n",
       "      <td>2.048969</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.221056</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.013642</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">Sphere</th>\n",
       "      <th rowspan=\"6\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.980354</td>\n",
       "      <td>3.082889</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.000285</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.028610</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.814091</td>\n",
       "      <td>5.070053</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.000145</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.016087</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>8.834389</td>\n",
       "      <td>9.977898</td>\n",
       "      <td>9.0</td>\n",
       "      <td>11.2</td>\n",
       "      <td>10.972066</td>\n",
       "      <td>11.0</td>\n",
       "      <td>9.149028</td>\n",
       "      <td>9.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>20.290503</td>\n",
       "      <td>32.557897</td>\n",
       "      <td>20.9</td>\n",
       "      <td>30.4</td>\n",
       "      <td>30.984633</td>\n",
       "      <td>31.0</td>\n",
       "      <td>22.246151</td>\n",
       "      <td>22.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>28.893689</td>\n",
       "      <td>52.645787</td>\n",
       "      <td>30.2</td>\n",
       "      <td>50.6</td>\n",
       "      <td>50.631222</td>\n",
       "      <td>51.0</td>\n",
       "      <td>32.016338</td>\n",
       "      <td>34.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>35.904561</td>\n",
       "      <td>71.800092</td>\n",
       "      <td>37.7</td>\n",
       "      <td>69.9</td>\n",
       "      <td>71.151596</td>\n",
       "      <td>71.0</td>\n",
       "      <td>39.312578</td>\n",
       "      <td>41.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Spiral</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.761830</td>\n",
       "      <td>1.522823</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.094092</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.963084</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.766377</td>\n",
       "      <td>1.526826</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.097328</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.013120</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">Uniform</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1000.0</th>\n",
       "      <th>10.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.453239</td>\n",
       "      <td>4.721269</td>\n",
       "      <td>4.8</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.966521</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.687491</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>8.081392</td>\n",
       "      <td>9.156182</td>\n",
       "      <td>8.1</td>\n",
       "      <td>9.8</td>\n",
       "      <td>9.872748</td>\n",
       "      <td>10.0</td>\n",
       "      <td>8.554030</td>\n",
       "      <td>8.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>19.048783</td>\n",
       "      <td>29.433654</td>\n",
       "      <td>19.9</td>\n",
       "      <td>30.1</td>\n",
       "      <td>29.824383</td>\n",
       "      <td>30.0</td>\n",
       "      <td>19.724822</td>\n",
       "      <td>20.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>27.195976</td>\n",
       "      <td>51.231300</td>\n",
       "      <td>28.7</td>\n",
       "      <td>50.9</td>\n",
       "      <td>49.812635</td>\n",
       "      <td>50.0</td>\n",
       "      <td>27.944971</td>\n",
       "      <td>29.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>33.930374</td>\n",
       "      <td>72.475144</td>\n",
       "      <td>35.9</td>\n",
       "      <td>70.9</td>\n",
       "      <td>70.671178</td>\n",
       "      <td>70.0</td>\n",
       "      <td>34.762951</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          MLE     GeoMLE  MIND  DANCo  \\\n",
       "Dataset        Num    Dim  RealDim                                      \n",
       "Affine         1000.0 5.0  5.0       4.436176   5.000000   4.7    5.0   \n",
       "                      10.0 10.0      8.092389  10.000000   8.1    9.5   \n",
       "                      20.0 20.0     14.110037  19.837311  14.5   19.3   \n",
       "Affine_3to5    1000.0 5.0  3.0       2.819162   3.125630   3.0    3.0   \n",
       "Cubic          1000.0 5.0  3.0       3.069715   3.120422   3.0    3.0   \n",
       "                      35.0 30.0     19.548624  30.146128  20.5   30.8   \n",
       "                      90.0 70.0     34.217167  73.697099  35.9   73.4   \n",
       "Helicoid       1000.0 3.0  2.0       2.086243   2.263129   2.0    2.0   \n",
       "                      13.0 2.0       2.079526   2.286545   2.0    2.0   \n",
       "Helix1d        1000.0 3.0  1.0       1.002463   1.399572   1.0    1.0   \n",
       "                      13.0 1.0       1.002080   1.403783   1.0    1.0   \n",
       "Helix2d        1000.0 3.0  2.0       2.741861   2.772936   2.9    3.0   \n",
       "                      13.0 2.0       2.743644   2.833216   2.9    3.0   \n",
       "Moebius        1000.0 3.0  2.0       1.985567   1.271459   2.0    2.0   \n",
       "Nonlinear      1000.0 8.0  4.0       4.076561   3.409430   4.0    5.7   \n",
       "                      36.0 6.0       6.683937   6.807444   6.0    8.1   \n",
       "                      64.0 8.0       9.210539  10.525456   9.0   10.5   \n",
       "                      72.0 12.0     13.437732  17.379927  13.6   17.5   \n",
       "Nonlinear_4to6 1000.0 6.0  4.0       3.825456   4.138118   4.0    5.0   \n",
       "Norm           1000.0 20.0 20.0     14.988604  20.000000  15.6   20.0   \n",
       "                      50.0 20.0     14.989692  22.124816  15.8   21.4   \n",
       "                           50.0     27.026298  50.000000  28.6   47.5   \n",
       "                      70.0 50.0     27.168977  53.282319  28.7   52.9   \n",
       "Paraboloid     1000.0 12.0 3.0       2.830874   2.999793   3.0    3.0   \n",
       "                      21.0 6.0       4.707588   6.011846   5.0    6.4   \n",
       "                      30.0 9.0       5.882575   9.163055   6.0    8.0   \n",
       "                      39.0 12.0      6.611216  12.152166   7.0    8.3   \n",
       "                      48.0 15.0      6.974033  13.636721   7.4    8.7   \n",
       "Roll           1000.0 3.0  2.0       1.954331   2.275757   2.0    2.0   \n",
       "                      13.0 2.0       1.947018   2.317430   2.0    2.0   \n",
       "Scurve         1000.0 3.0  2.0       1.946351   2.037241   2.0    2.0   \n",
       "                      13.0 2.0       1.948070   2.048969   2.0    2.0   \n",
       "Sphere         1000.0 5.0  3.0       2.980354   3.082889   3.0    3.0   \n",
       "                      10.0 5.0       4.814091   5.070053   5.0    6.0   \n",
       "                      15.0 10.0      8.834389   9.977898   9.0   11.2   \n",
       "                      35.0 30.0     20.290503  32.557897  20.9   30.4   \n",
       "                      55.0 50.0     28.893689  52.645787  30.2   50.6   \n",
       "                      75.0 70.0     35.904561  71.800092  37.7   69.9   \n",
       "Spiral         1000.0 3.0  1.0       1.761830   1.522823   1.0    1.0   \n",
       "                      13.0 1.0       1.766377   1.526826   1.0    1.0   \n",
       "Uniform        1000.0 10.0 5.0       4.453239   4.721269   4.8    5.0   \n",
       "                      15.0 10.0      8.081392   9.156182   8.1    9.8   \n",
       "                      35.0 30.0     19.048783  29.433654  19.9   30.1   \n",
       "                      55.0 50.0     27.195976  51.231300  28.7   50.9   \n",
       "                      75.0 70.0     33.930374  72.475144  35.9   70.9   \n",
       "\n",
       "                                          ESS   PCA         CD  Hein  \n",
       "Dataset        Num    Dim  RealDim                                    \n",
       "Affine         1000.0 5.0  5.0       4.961017   5.0   4.737116   4.8  \n",
       "                      10.0 10.0     10.020648  10.0   8.609721   8.5  \n",
       "                      20.0 20.0     20.012921  20.0  14.791872  15.2  \n",
       "Affine_3to5    1000.0 5.0  3.0       2.581466   3.0   2.917733   3.0  \n",
       "Cubic          1000.0 5.0  3.0       3.953560   4.0   3.058121   3.0  \n",
       "                      35.0 30.0     30.946966  31.0  20.579042  21.3  \n",
       "                      90.0 70.0     71.543753  71.0  34.738716  37.4  \n",
       "Helicoid       1000.0 3.0  2.0       1.839381   2.0   2.031638   2.0  \n",
       "                      13.0 2.0       1.833291   2.0   2.033630   2.0  \n",
       "Helix1d        1000.0 3.0  1.0       2.641996   3.0   1.024649   1.0  \n",
       "                      13.0 1.0       2.613153   3.0   1.006132   1.0  \n",
       "Helix2d        1000.0 3.0  2.0       2.550043   3.0   2.017534   2.0  \n",
       "                      13.0 2.0       2.547343   3.0   2.022974   2.0  \n",
       "Moebius        1000.0 3.0  2.0       2.314794   3.0   2.020746   2.0  \n",
       "Nonlinear      1000.0 8.0  4.0       7.942824   8.0   3.760750   4.0  \n",
       "                      36.0 6.0      11.961104  12.0   5.746971   5.9  \n",
       "                      64.0 8.0      16.027877  16.0   7.982724   7.9  \n",
       "                      72.0 12.0     23.923810  24.0  12.080877  11.9  \n",
       "Nonlinear_4to6 1000.0 6.0  4.0       3.144661   5.0   3.678472   4.0  \n",
       "Norm           1000.0 20.0 20.0     20.045761  20.0  13.565712  14.4  \n",
       "                      50.0 20.0     20.271046  20.0  13.780381  14.7  \n",
       "                           50.0     49.690825  50.0  24.035341  26.1  \n",
       "                      70.0 50.0     49.868543  50.0  24.444687  26.4  \n",
       "Paraboloid     1000.0 12.0 3.0       1.786439   1.0   2.196276   2.0  \n",
       "                      21.0 6.0       1.472586   1.0   2.561571   2.8  \n",
       "                      30.0 9.0       1.308293   1.0   2.766422   2.8  \n",
       "                      39.0 12.0      1.223703   1.0   3.025258   3.0  \n",
       "                      48.0 15.0      1.161489   1.0   3.327201   3.1  \n",
       "Roll           1000.0 3.0  2.0       2.952127   3.0   1.998514   2.0  \n",
       "                      13.0 2.0       2.974154   3.0   2.030448   2.0  \n",
       "Scurve         1000.0 3.0  2.0       2.182087   3.0   1.939938   2.0  \n",
       "                      13.0 2.0       2.221056   3.0   2.013642   2.0  \n",
       "Sphere         1000.0 5.0  3.0       4.000285   4.0   3.028610   3.0  \n",
       "                      10.0 5.0       6.000145   6.0   5.016087   5.0  \n",
       "                      15.0 10.0     10.972066  11.0   9.149028   9.2  \n",
       "                      35.0 30.0     30.984633  31.0  22.246151  22.6  \n",
       "                      55.0 50.0     50.631222  51.0  32.016338  34.0  \n",
       "                      75.0 70.0     71.151596  71.0  39.312578  41.2  \n",
       "Spiral         1000.0 3.0  1.0       2.094092   2.0   0.963084   1.0  \n",
       "                      13.0 1.0       2.097328   2.0   1.013120   1.0  \n",
       "Uniform        1000.0 10.0 5.0       4.966521   5.0   4.687491   5.0  \n",
       "                      15.0 10.0      9.872748  10.0   8.554030   8.6  \n",
       "                      35.0 30.0     29.824383  30.0  19.724822  20.3  \n",
       "                      55.0 50.0     49.812635  50.0  27.944971  29.6  \n",
       "                      75.0 70.0     70.671178  70.0  34.762951  37.0  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['Dataset', 'Num', 'Dim', 'RealDim']).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dfr = df.copy().round()\n",
    "names = dfr.columns.difference(['Dataset', 'Num', 'Dim', 'RealDim'])\n",
    "\n",
    "f, ax = plt.subplots(1, 1, figsize=(12, 8))\n",
    "nsteps = 5001\n",
    "max_tau = 40\n",
    "errors_total = abs(dfr[names].sub(dfr['RealDim'], axis=0).values)\n",
    "errors_total = np.max([errors_total, np.ones_like(errors_total)], axis=0)\n",
    "\n",
    "\n",
    "taus = np.linspace(1, max_tau, nsteps)        \n",
    "rhos = np.zeros((nsteps, len(names)))\n",
    "for cnt, tau in enumerate(taus):\n",
    "    rhos[cnt] = np.sum(errors_total <= tau, axis=0)\n",
    "\n",
    "for i in range(len(names)):\n",
    "    ax.plot(taus, rhos[:,i]/errors_total.shape[0], label=names[i])\n",
    "\n",
    "ax.set_title('Dolan-More Curve')\n",
    "ax.legend()\n",
    "ax.grid()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('result/result_dm.csv', index_label=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Nonunform samplings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()\n",
    "DG = DataGenerator()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['Helix1d', 'Helix2d', 'Helicoid', 'Spiral', 'Roll', 'Scurve', 'Star', 'Moebius', 'Sphere', 'Norm', 'Uniform', 'Cubic', 'Affine_3to5', 'Affine', 'Nonlinear_4to6', 'Nonlinear', 'Paraboloid'])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DG.dict_gen.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "steps = [\n",
    "    ('Sphere',  5,  3),\n",
    "    ('Sphere',  10, 5),\n",
    "    ('Sphere',  15, 10),\n",
    "    ('Sphere',  35, 30),\n",
    "    ('Sphere',  55, 50),\n",
    "    ('Sphere',  75, 70),\n",
    "    \n",
    "    ('Affine_3to5',  5,  3),\n",
    "    ('Affine',       5,  5),\n",
    "    ('Affine',       10, 10),\n",
    "    ('Affine',       20, 20),\n",
    "    \n",
    "    ('Nonlinear_4to6',  6,  4),\n",
    "    ('Nonlinear',       8,  4),\n",
    "    ('Nonlinear',       36, 6),\n",
    "    ('Nonlinear',       64, 8),\n",
    "    ('Nonlinear',       72, 12),\n",
    "    \n",
    "    ('Spiral',   3,  1),\n",
    "    ('Spiral',   13, 1),\n",
    "    ('Helix1d',  3,  1),\n",
    "    ('Helix1d',  13, 1),\n",
    "    ('Helix2d',  3,  2),\n",
    "    ('Helix2d',  13, 2),\n",
    "    ('Helicoid', 3,  2),\n",
    "    ('Helicoid', 13, 2),\n",
    "    ('Roll',     3,  2),\n",
    "    ('Roll',     13, 2),\n",
    "    ('Scurve',   3,  2),\n",
    "    ('Scurve',   13, 2),\n",
    "    ('Moebius',  3,  2),\n",
    "    \n",
    "    ('Norm', 10, 5),\n",
    "    ('Norm', 15, 10),\n",
    "    ('Norm', 35, 30),\n",
    "    ('Norm', 55, 50),\n",
    "    ('Norm', 75, 70),\n",
    "     \n",
    "#     ('Norm', 20, 20),\n",
    "#     ('Norm', 50, 20),\n",
    "#     ('Norm', 50, 50),\n",
    "#     ('Norm', 70, 50),\n",
    "    \n",
    "    ('Paraboloid', 12, 3),\n",
    "    ('Paraboloid', 21, 6),\n",
    "    ('Paraboloid', 30, 9),\n",
    "    ('Paraboloid', 39, 12),\n",
    "    ('Paraboloid', 48, 15),\n",
    "    \n",
    "#     ('Cubic', 5,  3),\n",
    "#     ('Cubic', 35, 30),\n",
    "#     ('Cubic', 90, 70),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_args(const_args, step):\n",
    "    args = {**const_args}\n",
    "    args['name'] = step[0]\n",
    "    args['D'] = step[1]\n",
    "    args['d'] = step[2]\n",
    "    assert len({'name', 'D', 'd', 'N', 'noise'} & set(args.keys())) == 5\n",
    "    return args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 380/380 [06:53<00:00,  1.26it/s]\n"
     ]
    }
   ],
   "source": [
    "n_repetition = 10\n",
    "args = {'N': 1000, 'noise': 0, 'type_sample': 'nonuniform'}\n",
    "df, df_time = run_test(run_one_test, steps, args, n_repetition)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>MLE</th>\n",
       "      <th>GeoMLE</th>\n",
       "      <th>MIND</th>\n",
       "      <th>DANCo</th>\n",
       "      <th>ESS</th>\n",
       "      <th>PCA</th>\n",
       "      <th>CD</th>\n",
       "      <th>Hein</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dataset</th>\n",
       "      <th>Num</th>\n",
       "      <th>Dim</th>\n",
       "      <th>RealDim</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Affine</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>5.020814</td>\n",
       "      <td>4.998816</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.968821</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.813513</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>8.990513</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>10.046645</td>\n",
       "      <td>10.0</td>\n",
       "      <td>8.437193</td>\n",
       "      <td>8.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20.0</th>\n",
       "      <th>20.0</th>\n",
       "      <td>15.107916</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>15.9</td>\n",
       "      <td>19.9</td>\n",
       "      <td>19.893082</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.587803</td>\n",
       "      <td>14.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Affine_3to5</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>3.063209</td>\n",
       "      <td>3.037318</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.1</td>\n",
       "      <td>2.622393</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.985264</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helicoid</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.099377</td>\n",
       "      <td>2.022600</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.595938</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.028393</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.096877</td>\n",
       "      <td>2.018808</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.591167</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.015797</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helix1d</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.069185</td>\n",
       "      <td>0.936972</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.878401</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.959202</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.045382</td>\n",
       "      <td>0.584697</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.857758</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.975039</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Helix2d</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.471712</td>\n",
       "      <td>1.549056</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.1</td>\n",
       "      <td>2.093870</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.015928</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.472843</td>\n",
       "      <td>1.806733</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.112224</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.964734</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moebius</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.062018</td>\n",
       "      <td>2.022438</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.805454</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.988716</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">Nonlinear</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">1000.0</th>\n",
       "      <th>8.0</th>\n",
       "      <th>4.0</th>\n",
       "      <td>4.081218</td>\n",
       "      <td>3.993467</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.195866</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.870462</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36.0</th>\n",
       "      <th>6.0</th>\n",
       "      <td>5.998835</td>\n",
       "      <td>6.423197</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.1</td>\n",
       "      <td>7.779084</td>\n",
       "      <td>12.0</td>\n",
       "      <td>5.654124</td>\n",
       "      <td>5.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64.0</th>\n",
       "      <th>8.0</th>\n",
       "      <td>7.792259</td>\n",
       "      <td>8.730836</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10.451486</td>\n",
       "      <td>16.0</td>\n",
       "      <td>7.233379</td>\n",
       "      <td>7.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72.0</th>\n",
       "      <th>12.0</th>\n",
       "      <td>10.855378</td>\n",
       "      <td>13.433242</td>\n",
       "      <td>11.0</td>\n",
       "      <td>13.2</td>\n",
       "      <td>15.576719</td>\n",
       "      <td>24.0</td>\n",
       "      <td>9.823021</td>\n",
       "      <td>10.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Nonlinear_4to6</th>\n",
       "      <th>1000.0</th>\n",
       "      <th>6.0</th>\n",
       "      <th>4.0</th>\n",
       "      <td>3.926262</td>\n",
       "      <td>4.070970</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.959484</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.757939</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">Norm</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1000.0</th>\n",
       "      <th>10.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.990759</td>\n",
       "      <td>5.396950</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.958178</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.717071</td>\n",
       "      <td>4.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>9.023456</td>\n",
       "      <td>10.704392</td>\n",
       "      <td>9.1</td>\n",
       "      <td>10.3</td>\n",
       "      <td>9.898405</td>\n",
       "      <td>10.0</td>\n",
       "      <td>8.358798</td>\n",
       "      <td>8.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>19.672006</td>\n",
       "      <td>32.112742</td>\n",
       "      <td>20.6</td>\n",
       "      <td>31.4</td>\n",
       "      <td>29.963357</td>\n",
       "      <td>30.0</td>\n",
       "      <td>17.885251</td>\n",
       "      <td>19.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>27.073239</td>\n",
       "      <td>51.709286</td>\n",
       "      <td>28.4</td>\n",
       "      <td>50.7</td>\n",
       "      <td>49.714747</td>\n",
       "      <td>50.0</td>\n",
       "      <td>24.072621</td>\n",
       "      <td>25.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>33.057989</td>\n",
       "      <td>72.317066</td>\n",
       "      <td>35.1</td>\n",
       "      <td>68.2</td>\n",
       "      <td>69.729688</td>\n",
       "      <td>70.0</td>\n",
       "      <td>29.217024</td>\n",
       "      <td>31.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">Paraboloid</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1000.0</th>\n",
       "      <th>12.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.832962</td>\n",
       "      <td>2.941758</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.785243</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.176015</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21.0</th>\n",
       "      <th>6.0</th>\n",
       "      <td>4.729833</td>\n",
       "      <td>5.812629</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.6</td>\n",
       "      <td>1.473727</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.608070</td>\n",
       "      <td>2.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30.0</th>\n",
       "      <th>9.0</th>\n",
       "      <td>5.931393</td>\n",
       "      <td>8.692435</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1.308759</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.939248</td>\n",
       "      <td>2.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39.0</th>\n",
       "      <th>12.0</th>\n",
       "      <td>6.644121</td>\n",
       "      <td>11.836172</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.2</td>\n",
       "      <td>1.223316</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.959303</td>\n",
       "      <td>2.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48.0</th>\n",
       "      <th>15.0</th>\n",
       "      <td>6.962873</td>\n",
       "      <td>14.016743</td>\n",
       "      <td>7.4</td>\n",
       "      <td>8.8</td>\n",
       "      <td>1.164491</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.223362</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Roll</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.033722</td>\n",
       "      <td>2.035524</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.196463</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.000428</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.048561</td>\n",
       "      <td>2.105231</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.179873</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.012574</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Scurve</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.033442</td>\n",
       "      <td>2.006502</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.743664</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.988450</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>2.0</th>\n",
       "      <td>2.040631</td>\n",
       "      <td>2.032976</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.742716</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.014277</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">Sphere</th>\n",
       "      <th rowspan=\"6\" valign=\"top\">1000.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>3.0</th>\n",
       "      <td>2.961716</td>\n",
       "      <td>3.019967</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.993347</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.828709</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <th>5.0</th>\n",
       "      <td>4.631092</td>\n",
       "      <td>4.740091</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.1</td>\n",
       "      <td>6.003999</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.736865</td>\n",
       "      <td>4.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <th>10.0</th>\n",
       "      <td>8.189919</td>\n",
       "      <td>8.526787</td>\n",
       "      <td>8.1</td>\n",
       "      <td>10.4</td>\n",
       "      <td>11.025315</td>\n",
       "      <td>11.0</td>\n",
       "      <td>8.392103</td>\n",
       "      <td>8.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <th>30.0</th>\n",
       "      <td>19.792762</td>\n",
       "      <td>29.582027</td>\n",
       "      <td>20.5</td>\n",
       "      <td>29.6</td>\n",
       "      <td>30.811456</td>\n",
       "      <td>31.0</td>\n",
       "      <td>21.138476</td>\n",
       "      <td>21.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55.0</th>\n",
       "      <th>50.0</th>\n",
       "      <td>28.389587</td>\n",
       "      <td>50.090070</td>\n",
       "      <td>29.7</td>\n",
       "      <td>49.5</td>\n",
       "      <td>50.971918</td>\n",
       "      <td>51.0</td>\n",
       "      <td>31.093704</td>\n",
       "      <td>31.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75.0</th>\n",
       "      <th>70.0</th>\n",
       "      <td>35.100652</td>\n",
       "      <td>67.677164</td>\n",
       "      <td>36.7</td>\n",
       "      <td>69.3</td>\n",
       "      <td>71.342877</td>\n",
       "      <td>71.0</td>\n",
       "      <td>38.185681</td>\n",
       "      <td>40.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Spiral</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">1000.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.547495</td>\n",
       "      <td>1.974704</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.010344</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.988890</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1.528277</td>\n",
       "      <td>1.952715</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.022800</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.982089</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          MLE     GeoMLE  MIND  DANCo  \\\n",
       "Dataset        Num    Dim  RealDim                                      \n",
       "Affine         1000.0 5.0  5.0       5.020814   4.998816   5.0    5.0   \n",
       "                      10.0 10.0      8.990513  10.000000   9.0   10.0   \n",
       "                      20.0 20.0     15.107916  20.000000  15.9   19.9   \n",
       "Affine_3to5    1000.0 5.0  3.0       3.063209   3.037318   3.0    3.1   \n",
       "Helicoid       1000.0 3.0  2.0       2.099377   2.022600   2.0    2.0   \n",
       "                      13.0 2.0       2.096877   2.018808   2.0    2.0   \n",
       "Helix1d        1000.0 3.0  1.0       1.069185   0.936972   1.0    1.0   \n",
       "                      13.0 1.0       1.045382   0.584697   1.0    1.0   \n",
       "Helix2d        1000.0 3.0  2.0       2.471712   1.549056   2.0    2.1   \n",
       "                      13.0 2.0       2.472843   1.806733   2.0    2.0   \n",
       "Moebius        1000.0 3.0  2.0       2.062018   2.022438   2.0    2.0   \n",
       "Nonlinear      1000.0 8.0  4.0       4.081218   3.993467   4.0    5.0   \n",
       "                      36.0 6.0       5.998835   6.423197   6.0    7.1   \n",
       "                      64.0 8.0       7.792259   8.730836   8.0    9.0   \n",
       "                      72.0 12.0     10.855378  13.433242  11.0   13.2   \n",
       "Nonlinear_4to6 1000.0 6.0  4.0       3.926262   4.070970   4.0    5.0   \n",
       "Norm           1000.0 10.0 5.0       4.990759   5.396950   5.0    6.0   \n",
       "                      15.0 10.0      9.023456  10.704392   9.1   10.3   \n",
       "                      35.0 30.0     19.672006  32.112742  20.6   31.4   \n",
       "                      55.0 50.0     27.073239  51.709286  28.4   50.7   \n",
       "                      75.0 70.0     33.057989  72.317066  35.1   68.2   \n",
       "Paraboloid     1000.0 12.0 3.0       2.832962   2.941758   3.0    3.0   \n",
       "                      21.0 6.0       4.729833   5.812629   5.0    6.6   \n",
       "                      30.0 9.0       5.931393   8.692435   6.0    8.0   \n",
       "                      39.0 12.0      6.644121  11.836172   7.0    8.2   \n",
       "                      48.0 15.0      6.962873  14.016743   7.4    8.8   \n",
       "Roll           1000.0 3.0  2.0       2.033722   2.035524   2.0    2.0   \n",
       "                      13.0 2.0       2.048561   2.105231   2.0    2.0   \n",
       "Scurve         1000.0 3.0  2.0       2.033442   2.006502   2.0    2.0   \n",
       "                      13.0 2.0       2.040631   2.032976   2.0    2.0   \n",
       "Sphere         1000.0 5.0  3.0       2.961716   3.019967   3.0    3.0   \n",
       "                      10.0 5.0       4.631092   4.740091   5.0    6.1   \n",
       "                      15.0 10.0      8.189919   8.526787   8.1   10.4   \n",
       "                      35.0 30.0     19.792762  29.582027  20.5   29.6   \n",
       "                      55.0 50.0     28.389587  50.090070  29.7   49.5   \n",
       "                      75.0 70.0     35.100652  67.677164  36.7   69.3   \n",
       "Spiral         1000.0 3.0  1.0       1.547495   1.974704   1.0    1.0   \n",
       "                      13.0 1.0       1.528277   1.952715   1.0    1.0   \n",
       "\n",
       "                                          ESS   PCA         CD  Hein  \n",
       "Dataset        Num    Dim  RealDim                                    \n",
       "Affine         1000.0 5.0  5.0       4.968821   5.0   4.813513   5.0  \n",
       "                      10.0 10.0     10.046645  10.0   8.437193   8.8  \n",
       "                      20.0 20.0     19.893082  20.0  13.587803  14.4  \n",
       "Affine_3to5    1000.0 5.0  3.0       2.622393   3.0   2.985264   3.0  \n",
       "Helicoid       1000.0 3.0  2.0       2.595938   3.0   2.028393   2.0  \n",
       "                      13.0 2.0       2.591167   3.0   2.015797   2.0  \n",
       "Helix1d        1000.0 3.0  1.0       2.878401   3.0   0.959202   1.0  \n",
       "                      13.0 1.0       2.857758   3.0   0.975039   1.0  \n",
       "Helix2d        1000.0 3.0  2.0       2.093870   2.0   2.015928   2.0  \n",
       "                      13.0 2.0       2.112224   2.0   1.964734   2.0  \n",
       "Moebius        1000.0 3.0  2.0       1.805454   2.0   1.988716   2.0  \n",
       "Nonlinear      1000.0 8.0  4.0       5.195866   8.0   3.870462   4.0  \n",
       "                      36.0 6.0       7.779084  12.0   5.654124   5.9  \n",
       "                      64.0 8.0      10.451486  16.0   7.233379   7.6  \n",
       "                      72.0 12.0     15.576719  24.0   9.823021  10.2  \n",
       "Nonlinear_4to6 1000.0 6.0  4.0       2.959484   4.0   3.757939   4.0  \n",
       "Norm           1000.0 10.0 5.0       4.958178   5.0   4.717071   4.9  \n",
       "                      15.0 10.0      9.898405  10.0   8.358798   8.9  \n",
       "                      35.0 30.0     29.963357  30.0  17.885251  19.2  \n",
       "                      55.0 50.0     49.714747  50.0  24.072621  25.9  \n",
       "                      75.0 70.0     69.729688  70.0  29.217024  31.9  \n",
       "Paraboloid     1000.0 12.0 3.0       1.785243   1.0   2.176015   2.0  \n",
       "                      21.0 6.0       1.473727   1.0   2.608070   2.6  \n",
       "                      30.0 9.0       1.308759   1.0   2.939248   2.9  \n",
       "                      39.0 12.0      1.223316   1.0   2.959303   2.9  \n",
       "                      48.0 15.0      1.164491   1.0   3.223362   3.0  \n",
       "Roll           1000.0 3.0  2.0       2.196463   3.0   2.000428   2.0  \n",
       "                      13.0 2.0       2.179873   3.0   2.012574   2.0  \n",
       "Scurve         1000.0 3.0  2.0       1.743664   3.0   1.988450   2.0  \n",
       "                      13.0 2.0       1.742716   3.0   2.014277   2.0  \n",
       "Sphere         1000.0 5.0  3.0       3.993347   4.0   2.828709   3.0  \n",
       "                      10.0 5.0       6.003999   6.0   4.736865   4.9  \n",
       "                      15.0 10.0     11.025315  11.0   8.392103   8.6  \n",
       "                      35.0 30.0     30.811456  31.0  21.138476  21.4  \n",
       "                      55.0 50.0     50.971918  51.0  31.093704  31.7  \n",
       "                      75.0 70.0     71.342877  71.0  38.185681  40.9  \n",
       "Spiral         1000.0 3.0  1.0       2.010344   2.0   0.988890   1.0  \n",
       "                      13.0 1.0       2.022800   2.0   0.982089   1.0  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['Dataset', 'Num', 'Dim', 'RealDim']).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dfr = df.copy().round()\n",
    "names = dfr.columns.difference(['Dataset', 'Num', 'Dim', 'RealDim'])\n",
    "\n",
    "f, ax = plt.subplots(1, 1, figsize=(12, 8))\n",
    "nsteps = 5001\n",
    "max_tau = 40\n",
    "errors_total = abs(dfr[names].sub(dfr['RealDim'], axis=0).values)\n",
    "errors_total = np.max([errors_total, np.ones_like(errors_total)], axis=0)\n",
    "\n",
    "\n",
    "taus = np.linspace(1, max_tau, nsteps)        \n",
    "rhos = np.zeros((nsteps, len(names)))\n",
    "for cnt, tau in enumerate(taus):\n",
    "    rhos[cnt] = np.sum(errors_total <= tau, axis=0)\n",
    "\n",
    "for i in range(len(names)):\n",
    "    ax.plot(taus, rhos[:,i]/errors_total.shape[0], label=names[i])\n",
    "\n",
    "ax.set_title('Dolan-More Curve')\n",
    "ax.legend()\n",
    "ax.grid()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('result/result_nonuniform_dm.csv', index_label=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dependece on noise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()\n",
    "DG = DataGenerator()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['Helix1d', 'Helix2d', 'Helicoid', 'Spiral', 'Roll', 'Scurve', 'Star', 'Moebius', 'Sphere', 'Norm', 'Uniform', 'Cubic', 'Affine_3to5', 'Affine', 'Nonlinear_4to6', 'Nonlinear', 'Paraboloid'])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DG.dict_gen.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "steps = [\n",
    "    ('Sphere',  5,  3),\n",
    "    ('Sphere',  10, 5),\n",
    "    ('Sphere',  15, 10),\n",
    "    ('Sphere',  35, 30),\n",
    "    ('Sphere',  55, 50),\n",
    "    ('Sphere',  75, 70),\n",
    "    \n",
    "    ('Affine_3to5',  5,  3),\n",
    "    ('Affine',       5,  5),\n",
    "    ('Affine',       10, 10),\n",
    "    ('Affine',       20, 20),\n",
    "    \n",
    "    ('Nonlinear_4to6',  6,  4),\n",
    "    ('Nonlinear',       8,  4),\n",
    "    ('Nonlinear',       36, 6),\n",
    "    ('Nonlinear',       64, 8),\n",
    "    ('Nonlinear',       72, 12),\n",
    "    \n",
    "    ('Spiral',   3,  1),\n",
    "    ('Spiral',   13, 1),\n",
    "    ('Helix1d',  3,  1),\n",
    "    ('Helix1d',  13, 1),\n",
    "    ('Helix2d',  3,  2),\n",
    "    ('Helix2d',  13, 2),\n",
    "    ('Helicoid', 3,  2),\n",
    "    ('Helicoid', 13, 2),\n",
    "    ('Roll',     3,  2),\n",
    "    ('Roll',     13, 2),\n",
    "    ('Scurve',   3,  2),\n",
    "    ('Scurve',   13, 2),\n",
    "    ('Moebius',  3,  2),\n",
    "    \n",
    "    ('Uniform', 10, 5),\n",
    "    ('Uniform', 15, 10),\n",
    "    ('Uniform', 35, 30),\n",
    "    ('Uniform', 55, 50),\n",
    "    ('Uniform', 75, 70),\n",
    "     \n",
    "    ('Norm', 20, 20),\n",
    "    ('Norm', 50, 20),\n",
    "    ('Norm', 50, 50),\n",
    "    ('Norm', 70, 50),\n",
    "    \n",
    "    ('Paraboloid', 12, 3),\n",
    "    ('Paraboloid', 21, 6),\n",
    "    ('Paraboloid', 30, 9),\n",
    "    ('Paraboloid', 39, 12),\n",
    "    ('Paraboloid', 48, 15),\n",
    "    \n",
    "    ('Cubic', 5,  3),\n",
    "    ('Cubic', 35, 30),\n",
    "    ('Cubic', 90, 70),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_args(const_args, step):\n",
    "    args = {}\n",
    "    args['N'] = const_args['N']\n",
    "    args['noise'] = const_args['noise']\n",
    "    args['type_sample'] = const_args['type_sample']\n",
    "    args['name'] = step[0]\n",
    "    args['D'] = step[1]\n",
    "    args['d'] = step[2]\n",
    "    return args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "noise_steps = [0, 0.01, 0.02, 0.03, 0.04, 0.05]                     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 225/225 [04:27<00:00,  1.43s/it]\n",
      "100%|██████████| 225/225 [04:27<00:00,  1.15s/it]\n",
      "100%|██████████| 225/225 [04:27<00:00,  1.28s/it]\n",
      "100%|██████████| 225/225 [04:28<00:00,  1.55s/it]\n",
      "100%|██████████| 225/225 [04:27<00:00,  1.11s/it]\n",
      "100%|██████████| 225/225 [04:27<00:00,  1.03s/it]\n"
     ]
    }
   ],
   "source": [
    "names = DE.names\n",
    "n_repetition = 5\n",
    "\n",
    "mpe = []\n",
    "std = []\n",
    "for noise in noise_steps:\n",
    "    args = {'N': 1000, 'noise': noise, 'type_sample': 'uniform'}\n",
    "    df, _ = run_test(run_one_test, steps, args, n_repetition, 12)\n",
    "    dfr = df.round()\n",
    "    mpe_i = abs(dfr[names].sub(dfr['RealDim'], axis=0)).mul(dfr['RealDim']**-1, axis=0).values\n",
    "    mpe.append(mpe_i.mean(axis=0))\n",
    "    std.append(mpe_i.std(axis=0)/np.sqrt(mpe_i.shape[0]))\n",
    "\n",
    "mpe = np.array(mpe)\n",
    "std = np.array(std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(1, 1, figsize=(12, 8))\n",
    "\n",
    "for mpe_i, std_i, name in zip(mpe.T, std.T, names):\n",
    "    ax.plot(noise_steps, mpe_i, '-*', label=name)\n",
    "    ax.fill_between(noise_steps, mpe_i - std_i, mpe_i + std_i, alpha=0.2)\n",
    "    \n",
    "ax.legend()\n",
    "ax.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "result = pd.DataFrame(mpe, columns=names)\n",
    "result['noise'] = noise_steps\n",
    "result.to_csv('result/result_noise_mpe.csv', index_label=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "result = pd.DataFrame(std, columns=names)\n",
    "result['noise'] = noise_steps\n",
    "result.to_csv('result/result_noise_std.csv', index_label=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Time dependance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()\n",
    "DG = DataGenerator()\n",
    "DG.dict_gen.keys()\n",
    "\n",
    "steps = np.arange(5, 101, 5).tolist()\n",
    "args = {'name': 'Sphere',\n",
    "        'N': 1000,\n",
    "        'noise': 0}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_args(const_args, step):\n",
    "    args = {**const_args}\n",
    "    args['d'] = step\n",
    "    args['D'] = step + 1\n",
    "    assert len({'name', 'D', 'd', 'N', 'noise'} & set(args.keys())) == 5\n",
    "    return args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 100/100 [03:14<00:00,  2.68s/it]\n"
     ]
    }
   ],
   "source": [
    "names = DE.names\n",
    "n_repetition = 5\n",
    "\n",
    "df = run_test(run_one_test, steps, args, n_repetition, return_time=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_mean = df.groupby(['Dataset', 'Num', 'Dim', 'RealDim']).mean().reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(1, 1, figsize=(12, 8))\n",
    "\n",
    "for name in names:\n",
    "    ax.plot(df_mean['RealDim'], df_mean[name], label=name)\n",
    "    \n",
    "plt.yscale('log')\n",
    "ax.legend()\n",
    "ax.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('result/result_time_dim.csv', index_label=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Num"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()\n",
    "DG = DataGenerator()\n",
    "DG.dict_gen.keys()\n",
    "\n",
    "steps = np.concatenate((np.arange(100, 1000, 200), \n",
    "                          np.arange(1000, 2000, 250), \n",
    "                          np.arange(2000, 10001, 500))).tolist()  \n",
    "args = {'name': 'Sphere',\n",
    "        'D': 11,\n",
    "        'd': 10,\n",
    "        'noise': 0}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_args(const_args, step):\n",
    "    args = {**const_args}\n",
    "    args['N'] = step\n",
    "    assert len({'name', 'D', 'd', 'N', 'noise'} & set(args.keys())) == 5\n",
    "    return args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 95%|█████████▌| 124/130 [06:18<00:14,  2.48s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'name': 'Sphere', 'D': 11, 'd': 10, 'noise': 0, 'N': 8000}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 130/130 [06:33<00:00,  2.89s/it]\n"
     ]
    }
   ],
   "source": [
    "names = DE.names\n",
    "n_repetition = 5\n",
    "\n",
    "df = run_test(run_one_test, steps, args, n_repetition, return_time=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_mean = df.groupby(['Dataset', 'Num', 'Dim', 'RealDim']).mean().reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAssAAAHXCAYAAABK9BemAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOzdd5gkZ3nv/W91qs45TJ4Ns3kVVlpJBAErsLAEyAgbMNiH18YgouNxIjj79TGc1wFzgINlwDY+BmxjCyMMBmxYwTFJ0gIKK23enTw93dM5d1e9f1SH6Qm7K+3OTM/O/bmuvqq7uqamaluz89O99/M8iq7rCCGEEEIIIZYzbfQFCCGEEEII0askLAshhBBCCLEKCctCCCGEEEKsQsKyEEIIIYQQq5CwLIQQQgghxCokLAshhBBCCLEKCctCCCGEEEKsQsKyEEIIIYQQq7Cs1zdSFOVe4OWAF/i4rutfWa/vLYQQQgghxLNxRZVlRVE+oShKXFGUJ5bsv0tRlBOKopxWFOVdALquf07X9fuAtwE/eSXfVwghhBBCiPVwpW0YfwPctXiHoihm4MPA3cB+4PWKouxfdMhvNd8XQgghhBCip11RG4au699QFGXbkt23Aqd1XT8LoCjKZ4BXKoryFPA+4Eu6rh9b7ZyKorwFeAuAw+G4eXh4+Eou8aI0TcNkkrbtrUA+661DPuutQz7rrUM+661joz7rkydPJnRdj6z03lr0LA8CE4teTwK3Ab8A/AjgUxRlTNf1j670xbqu3w/cD3D48GH9kUceWYNLNBw9epQjR46s2flF75DPeuuQz3rrkM9665DPeuvYqM9aUZQLq723bgP8dF3/IPDB9fp+QgghhBBCXKm1qHNPAYt7J4aa+4QQQgghhNhU1iIsPwzsUhRlu6IoNuB1wOefyQkURblHUZT7M5nMGlyeEEIIIYQQl+dKp477NPBtYI+iKJOKorxJ1/U68PPAl4GngH/Udf3JZ3JeXdcf1HX9LT6f70ouTwghhBBCiCtypbNhvH6V/V8Evngl5xZCCCGEEGKjyTwsQgghhBBCrKInw7L0LAshhBBCiF7Qk2FZepaFEEIIIUQv6MmwLIQQQgghRC+QsCyEEEIIIcQqJCwLIYQQQgixip4MyzLATwghhBBC9IKeDMsywE8IIYQQQvSCngzLQgghhBBC9AIJy0IIIYQQQqxCwrIQQgghhBCr6MmwLAP8hBBCCCG2mFoZSy2/0VexjGWjL2Aluq4/CDx4+PDh+zb6WoQQQgghxDOg61ArQjEJxQVjW0p1XpcWVn6vVmR/4BDc+YqNvoMuPRmWhRBCCCFED9B1qGSbwbYVdFshd5XQW1yARmX1c9r94AyCMwSefogdMJ47AkzP1Qiu391dFgnLQgghhBDXukYdyhkj0JbTUEo3t6lFzxftWxyItfrK51RM4Ag0g24Q/CMwcKPx3BnqBGJHsPPc7gfz6vEzcfTo2tz/FZCwLIQQQgixGWiNVQJvepV9i55Xcxc/t9VpBFlHABx+CI+B49ZO6F0WgAPG8aaeHP52VUlYFkIIIYTYSJU8pMcXPS5AZrJZ9W2F4AxULjHxgcXeDLzN0Osbgr6D3ftaz7v2+cCirs+9bkI9GZYVRbkHuGdsbGyjL0UIIYQQ4sqsFIYXvy4tdB9vVo2g6wyBpw+i+5aE3EB34G1trY6Nub9rXE+GZZkNQwghhBCbRiUPmYnVw3Ax2X28WTX6e/0jMHCo89w/amxdkS3R3rBZ9GRYFkIIIYToGVcShvtvlDC8yUlYFkIIIcS1SdehXjbCbiUL1bzxvJqHSq6zXbav+bqchdy0hOEtTsKyEEIIIXpPow6FecjPEVj4ATzVDLWVnDGzQzvg5juvu8Juc5/euLzvZ3WCzQ2qB1Q32DzgG4ShmxcFYQnDW5GEZSGEEEKsD103ZnfIzzUf8eZjbvm2mAR0AG4AeGzJuUyWTri1uY2Aq3rA2w+qt7Nv6TE2t/F++3lzazKv8x/G1qFXq2jFIlqhYGwXP1+ytedycOTIRl9yFwnLQgghhLgylTwUVgq+S0NwHLTa8q83q+CJgTsGgW0wfKvx3B0Fd4zvn7jAodte0Kn4qm5jmjRFWfdbvdbpum4E10IRrbh4e+mgqxeLNAqFzrZgHK/XVvjMV6Io2A8cWNsbfBYkLAshhBBiOa0BhQTkZoywm5tdPQTXCsu/XjEZ7QrNwEt0f+d5e9t8rnovGnwzc0eh//q1u9drlK7r6MUi9VSaRipFI50ytqkU9VSKRmt/8716Kk0jnYb6Kiv2LaGoKiaXC5PTaTxcLsxuD6ZYX/u1sXVicrqM7ZLjF28Vh4OHHnpojf9UnrmeDMsyz7IQQgixRrSG0Qucm4Hc3KIwvOR1Pr5yv6/d3wm5gzcvCb+LQrAzKK0NV5lWLtNIpy8ReLv36dXqyiczmzH7/ZgDfiz+ALZt23EcCmAOBDB7vUaAdbm6g66zO+wqlp6MkVddT96lzLMshBBCPEOtAXGrhd/W60IcdG3517si4O4zFsHoOwiefiP0evqNfa0wLCu9PSO6pqGXSmilUqeNoVhEK5aM9obma71UarY8dFodWsG4nk7RSGfQi8WVv4miYPZ6jaAbCGAdGMB+YD+W5muzv7kN+Nv7TB4PigxSvCw9GZaFEEII0dSoNVseZi9eCS7MrxCCFSMEe5qht+/6Zvhtvm6FY3cUzNYNub1epOs6jVSK2tQ0Wi67JOAuCrQlo0/XCLfF5YG4VFo94K5CabUoOJ1GuI1EUHfvbgfhduD1+zv7vN4tU+XdCPInK4QQQmyEerXT/5ubafYELw7Bs0ZALiRozQrRoXRaHjz9MHDjkkpwc+uKSAheQScMT3U9qu3n0+il0kXPodjtRqh1OBb14DoxB4Od1633XZ3XisNhtDW0enkXfb1it0u1twdJWBZCCCGuplq5UwXOzxqhd2kQzs8uX+gCmoPioka11zcIgzctqQQvDsHyK3w1zyYMm3w+rIMDqNu3437+7VgHB7EODmD2+TphdlEAVszSj71VyE+aEEIIcbmKC5A8DZnJzgwRrQpwqyWinF7+dSZLZ+BbYBuM3NZpgWj1A3v6wRWWQXGX4aqF4aFBYzswgNnj2aC7Eb1OwrIQQgixWKMGC+cgeQoSpzrbxCkoLXQfa7J2Am9oJ2x7fjP8Lg7CfeAM9dyKb635dBvJJPXkAo2FJPVEkvpCkkZywdgmktQXFmgkkzTyeRSzGcViMfpjrdb2c8ViQbFawLLCPvPSYywoi49bus/aPP+ifc4f/pCZhx5qB+Ha1BR6udx1PxKGxVqRsCyEEGLr0XWjF3ilQJw63z1lmisK4V2w7x4I74bQGPiHjUqwI9BTC2PotZoxddjCAvVksjsIJxeoJxOdIJxcQK9UVjyPyefDEgxiDgVRx8aw3HYrJrcHvVGHeh29Vkevtx61ZftoGK+1Urmzr15b8nV1qNW6XzdWXpraA2RbYXjHdty3SxgW60fCshBCiGtXvQILZ5tB+KTRQtEKx+VM5zizaoTgvoNw4FVGOA7tMqrFDv+GXb5eq9HIZBZNIZbuCrvtQNyq/qZXaAEBsFqxhELNABxCHRvDHApiCYaMbSiMJWS8ZwkEUGy29b3RJl3TjOC9+FGr861jj/Kiu+/ekGsSoifDsixKIoQQ4rLputE/nDjZDMKLAnF6vHs6Nc8AhMfg4Ks7gTg8Br7hNe8V1kolI/S2FpVIp43wm04bi0iklzxSKbR8ftXzdVV/d+7EctutmIMhI/QGQ1jCIczBIJZQyJhTt4cq4KtRTCaw2ZaFdd3h2KArEqJHw7IsSiKEEGKZWrlZJTZC8b7j34STvweJ01DNdY6zOo2K8MBNcP1PdgJxaAzUK/+nel3X0XK5rtDbetTb+zLLgvFqLQ+AsUxwa+5cvx/b6Gh7dbXWvtbDEg5vaPVXiK2mJ8OyEEKILUrXjSnVEic7leJWC0X6QleV2KdGYPAg3Pj6ZiBuPjwDz2owXSsE1+Nx6vE4tXic+ly8/boej1Obj1OfT0CttvJJTKbOSmp+P9b+fuz79i0Kwr5O6G3t8/kk+ArRwyQsCyGEWH+NuhF+26H4ZCcUl1Kd4yx2IwgPHDKqxOFdzUF2O/nOtx7myJEjl/XttGKxE4Dj84sC8FzXvqUzLACYPB4s0SiWaATXLbdgiUYxB0Pt8GtZVBE2eb2yqIQQ1xgJy0IIIdZOOWO0SSwNxQtnQVtUnXVFjRC8/15jG95tBGPf8EWrxFq1uiT8xqnPN6vAi6rCK/X+KnY7llgUaySK4+BBIxDHYliiEazRqPE6EsHkdK7Fn4wQYpOQsCyEEOLKaBpkp7qrw63n+dnOcSYLBHcYQXjvy9qhWPdvR2tYaORyNDIZox/4bBbthw/TyPwnjVwWLZujkc0azzNZGrkckdlZThQKy6/HasUaiWCJRlHHxnA973lYolGssWYAbj5MbvemGPQmhNhYEpaFEEJcmq4bq9ctnIHkmeb2NCRPoyfOoJdLNKoKWs1EAw8N+yCa7QAN0+008KBpdholBe1Ensb3sjRyj6FlvkkjlzOqvrq++vc2mTB7PJi8XsxeLyavBzUWIz0wwMihGztV4FjMaJHw+aQVQghx1UhYFkII0VHONMPw2WYYPoM+f5raxDlqyQLVvIVawWxsy05qRRONcgAaS+ciXmg+DCan0wi7Hg8mnxdrXx/m3bvb+8w+LyaPt7n1GIPkvF5MXi8ml2vFCvDpo0eJXGbPshBCPFsSloUQYqupFpph2KgO68kzaFOnqI5foJbIGkE4b6ZasFArqtTygO4AjLluFavFWDVt7wj2/n7MPh8mrwez14fZ26kAG8HYh9ntRrFaN/SWhRDi2ZKwLIQQ16J6BRbOtdsl9LlT1C6cojo+Tm0+vSgQm6kVrGhVBeNXQhAAc8CHbXgEx8go3uEhbEPDWIeHsA0PY4lGUcxru4CHEEL0CgnLQgixWbWmX0ueQU+cojHxNLVzJ6lOTlKbSxtBuBmKayUz6K1WBr9RHe6PYj24HefItnYQtg4NYxsaxORybeitCSFEr5CwLIQQvay1lHPyNPr8KeqnH6N69hTViQlqcymqOVO7SqzVFw9q82L2u7EN9uO4ZQe+0Z1Yh4exDQ9hHR7GEonIIDghhLgMEpaFEKIXNAfW6XMnqZ36AdUzJ6hOTFKbTVLN6FTzRpVY1zoD3RSLB2s0gG3PEM7tu7Bt39UJxIODMj+wEEJcBRKWhRBivdQrkDqPNv0UtaePNQPxBNWZJLVU1agQFxe3S4Bic2KLBVAPDuHZsQvr2H5s27YZvcN9fdI7LIQQa6wnw7KiKPcA94yNjW30pQghxDPTXKCjMf441aePUTv9dDMQJ6gtlKnmzNRL3QHX5LBhi8Vw7BrEu2MXtt0HsW3fiW1kBHM4LAtnCCHEBurJsKzr+oPAg4cPH75vo69FCCFWpOs0pk5Q/cE3qD51jOrZ01Sn56kli1RzCo1KdyA2u63YYkO49g1g2zGGddd12HbtwzY8jNm/dI5iIYQQvaInw7IQQvQKrVw2BtQ99l9Un/4h1bNnqE7HqSbKNMrdFV+Lz4ot2o/nugFs23di3XM9tj03YBseltklhBBik5KwLITY8vRajdrUFNXz56k+/RiVE49RO3+OynSCeqbadazZoaGG7LivH0bdvh3b3huxXf88rGP7MKnqBt2BEEKItSJhWQixJeiaRn1ujuqFC0YoPnOa6qnjxuu5FGh6+1iTVcPmqeOMqtgO9WHbMYZt3yFsN7wA89ABkCnXhBBiy5CwLIS4Zui6jpLLUTx2jOr5Zig+f97oJx6fQK/W2scqZh2bp47qruPZB7ahPmw7d2PbfzPmsVtQ+g6A3beBdyOEEKIXSFgWQmxKjWyWyunTVE6epHLyFJVTxiOaTnOhdZAJbG4Nm7uKa3sdm6eOLRbENrYHy9iNKAPXQew6CG4Hk0zBJoQQYjkJy0KInqaVy1TOnGmH4crJU1ROnqQ+N9c+xqSaUAPgCeWxba8ZFWO/Gev23UYg7rsOYgeMhyOwgXcjhBBis5GwLIToCXq9TnV8fEml+CTV8Qlj7mJAMSvYAgpOVwH1hhqqr4bd38AyNIoS3QfRfTyZ0Dlwx09CcAeY5a84IYQQV0Z+kwgh1pWu69SnpykvqRRXz55Br9WNgxSw+RRUTwnvvhqqv4bqb2AbHkbp2w+RvRDdB5E9ENoFVnv7/PNHj0Jk98bcnBBCiGuOhGUhxJqpLywsqxRXTp5EK5bax1jcCqqnjGunUSlWfXXUbUOY+vdDdK8RjCN7Iby7KxQLIYQQ60HCshDiimmlUnuwXfnECSonTlI5+TSNVKZ9jNmuoHor+AarRiD21VFHBzAP7zcqxJF9RjgO7warYwPvRgghhOiQsCyEuGy6plGbmDAC8clTVE6cMFooxsdBN+YpViwKqq+K219F3VbH7qujjvZhHtnb7ituV4ptzg2+IyGEEOLiJCwLIVZUT6WaFeKTlE+eaLdS6KVmC4UCtqCK6i3jPZBD9dexh8xY99yAMnQzRA90KsU2WepZCCHE5iRhWYgtTqtWqZ4502yhMMJx5cQJ6vPz7WPMXidqzIl/n4LdmkP1llF9dUzhERi+HYZuheFbIHYQzNYNvBshhBDi6pKwLMQW0Z6FohWImxXj6rnz0GgAoNhs2Eb6cO0KoB5SUc0T2G1zmO0aitUBgzfB0Ctg6BYYvhXc0Y29KSGEEGKNSVgW4hrUyOU6gbjVX3zyJFo+3z7GOjiIumMEzw1D2F0FVNN5bOWnUDhvHOAflaqxEEKILU/CshCbmF6vU71wgcqJE10tFLXp6fYxJo8Hdc9ufC9/GWrMjurKoXIec+IY5B42DtIc0HcIht9hhOOhW8AT26C7EkIIIXqHhGUhNol6IrFoWrZmC8XpM+jVqnGAxYK6fRuOQ4fwv+51qINB7K40luJJlKmHYebfIV4zjvWPwrZm1XjosLEctFSNhRBCiGUkLAvRY7RymcrpM+1p2conjYDcWFhoH2OJRFD37MH1356Lfc9u1D17sEU9mKa/A2cfgnMfgofPNw92wMAheK5UjYUQQohnat3CsqIoO4D3Aj5d11+9Xt9XiF6l6zq1qal260SrjaJ6/jxoGgCK3Y66axfuO45g37MHdfce1D27sQQCUM7ChW/BuYfga++H+JPGiVUvjD4fbn0rjDxHqsZCCCHEFbiisKwoyieAVwBxXdcPLtp/F/AXgBn4mK7r79N1/SzwJkVRPnsl31OIzag14K7crBa3Wim0QqF9jHV4GHXPbrx33YW6ezfqnt3YRkZQzGbjgFoZJr8Hxz5kBOSpY6A3wKwaofjFvw07jkD/jWCWfzQSQgghroYr/Y36N8CHgE+2diiKYgY+DNwJTAIPK4ryeV3Xj1/h9xJiU9DrdcpPPUXxkUcpPvoI5ePHqU/PtN83eb3Yd+/Gd++9qLt3Y9+zG9vYLszuJQt3NOow8wM4d9RorZj4LtTLoJiNKdxu/xXY/kIYvg2s9vW9SSGEEGKLuKKwrOv6NxRF2bZk963A6WYlGUVRPgO8EpCwLK5JWqVC+fHHKT7yCMWHH6H0/e+jFYsAWEdGcB66CfV1e4ze4t27sfT1oSjK8hPpOsw/3ew5fgjO/xdUMsZ70QNw+OeMcDz6fLB71/EOhRBCiK1L0XX9yk5ghOUvtNowFEV5NXCXrutvbr5+A3Ab8LvAH2FUnD+m6/ofr3K+twBvAYjFYjd/5jOfuaLru5h8Po/b7V6z84vecTU/a6Vcxnr2LNZTp7CdOo31/HmUeh2A2uAgtbExqrvGqI2Nofn9Fz2XvTSHP/0YgZTxsNXSAJTsfaQC15MKXE/afx0128XPIzrk53rrkM9665DPeuvYqM/6jjvueFTX9cMrvbdujY26rieBt13GcfcD9wMcPnxYP3LkyJpd09GjR1nL84vecSWfdT2VonTsGMWHH6H46KOUjx83Vrwzm7EfOIDzDW/AecthnDfdhPkS4Zj8vFE1PveQUUFOXzD2u2Ow907Y/iLY/kIcgVEcwMCzuuKtTX6utw75rLcO+ay3jl78rNciLE8Bw4teDzX3CbEp1ObiFB95mNKjj1J8+BEqp04BxlLQjuuvJ/SW+3AePozzxhsxuVwXP1k5Cxf+q9NaEW92I6k+Y57j577TCMiRPbBSa4YQQgghNtRahOWHgV2KomzHCMmvA37qmZxAUZR7gHvGxsbW4PKE6NB1ndrkpFE1fsR41MbHATA5nThuugnvy1+O85bD2A8exKSqlzohzD0Bp74Kp//DGJSn1Y25jkeeA9e9Bna8yJixwmRehzsUQgghxJW40qnjPg0cAcKKokwCv6vr+scVRfl54MsYU8d9Qtf1J5/JeXVdfxB48PDhw/ddyfUJsZSuaVTPnGkPxis+8gj1eBwAs8+H45bDBF7/epyHD2PftxfFchk/IuUMnD0Kp74Cp/8Tcs2ZL/qug+f9Iux8MQzfCpZLBG0hhBBC9JwrnQ3j9avs/yLwxSs5txBXS+X0aZz/8R9MfPazlB55lEbaGERniUaNdopbDuM8fBjbzp0oJtOlT7ha9Vj1wc47YNedMPYj4Olb4zsTQgghxFqTlQvENak2O0v23/6NzINfoPL003iAysgI7he/uB2QrUNDK0/htpJyBs58HU5/dXn1+Pm/BGN3GstIy2IgQgghxDWlJ3+zS8+yeDYamQzZL3+Z7Bf+jeLDD4OuY7/+emLveTdPeDy88FWvuvyTtavHX4FTzeqx3pDqsRBCCLHF9GRYlp5lcbm0cpn80aNkvvAFCg99A71WwzY6Svid78T3ipdj27bNOO7o0UufbNXq8fVw+y9L9VgIIYTYguS3vth09EaD4ne/S+bBL5D76lfR8nnMkTCBn3o93lfcg/3ggctrr9B1mH3cCMeLq8d2nzEob+xOGHuJVI+FEEKIdZBPZKmlSht9GctIWBabgq7rlJ94kuwXHiTzxS/SmE9gcrnwvPSl+O55Bc7bbkMxX8ZUbKU0nP26EY5P/wfkZ439rerxrpfC4GGpHgshhBBrpFqpMnt6kpmzk8Sn54inEiTLaYpU6DMHufNVd2/0JXaRRCB6Wm1mhsy/fp7M5z5H9fx5FKsV14teiO8V9+A+8iJMdvulTxJ/ipELn4VPvH+V6vGPgCe29jcjhBBCbCGNRoPkXIKZUxPMjk8TT8yTyC+QqReg+Q/AZt1EwOJmxNdPNBylTHVjL3oFPRmWZYDf1qaVSuT+4z/JPPAAhW9/G3Qdx+Gb6XvTz+F96Usx+3yXPkn8aXjyATj+OZh/mh0A/TfA7b9iDM6T6rEQQghxVei6TjabZfbcFDPnppibnSWRTrJQydJAax4EXsVJ2O5nb3Qnsf4Y/TuHie4awGzr/D4+ejljjNZZT6YFGeC39ei6Tun73yfzwANkv/gltEIB6+Ag4be/Hd+9r8Q2MnLpk8yfNALykw/A/FOAYiwpfet9fGshyPN+9MfX/D6EEEKIa1mpVGJudpaZM1PMTc4QT8yTLKSo6LX2MU5dJWj2cJ1vjFg0St/oIH27h7BH3Jc/ZWsP6cmwLLaO2vQ0mc9/nvQDD1C7MI7idOJ96UvxvepVOG85fOlFQhKn4MnPGQE5/iSgwOjz4GV/Avt+rN1eUe3B/1MVQgghelWtVmN+fp65yRlmzk8Tn5tjPpOkUO8MwLPqZoK42ekYJBoMExvop29sEN+2MCb7tRMxr507EZuGViqR++pXST/wAMXvfBd0HeettxJ+29vxvvROTC7XxU+QOA3HHzBC8twTgAIjz4W7/6cRkL3963IfQgghRC/TNI1KpdL1KJfLlAulzqNYplwuUymXjWOqVfKlPOlSFr15HpOu4Ndd9JsDRHw7ifbF6B8dILgjhjXiQjFvvmrxMyFhWawLXdcpHTtG+oEHyH3p3402i6EhYz7ke1+JbWjo4idYONtpsZh93Ng3/By46/2w/8fAO7D2NyGEEEKsk3q9boTYSoVyqUw5X6RcKFEqlKgUy5SLJcqlVsCtUK5UqNaqVOrGo9qoUdPrl/w+ig42LFixYNMt2LDg1VV2OHYRCUWIDfUT2zmAOujB7LGtw533np4MyzLA79qh6zrFb3+b+Af+gvJjjxltFnfdhf9V9+K4+eaLt1k06nDy3+HhjxnTvQEM3Qo/+sdGQPZdImALIYQQPaharZLNZEnHF8jEF0gn02QzGbK5HLlSnnytSFm79KwQZt2EbVHItWLBa1JRzR5sViuqxYZqs6HaVFSbHbtdRbXbsTuNh+pyojpVTHYLJtWMopoxqRZMDguK9RJtkFtIT4ZlGeB3bSgeO8b8n3+A4sMPY+nvp+/3fx/fK15+6TaLfByOfRIe+WvIToJ3EO54L9zwevAPr8/FCyGEEM9CtVolm86QmkuSiafJJFNkM1my+Sy5coFctdA1GK5F1a24dBW3xUlMDeBxurCrdlTVjqqq2B2dkGt3O7G7HFgcNkz2Vsg1g8W0aQbQ6bpOKVcjNVMgNVckNVsgPVskX9PgyEZfXbeeDMticys9+STzf/EXFL7xTczhMLH3vhf/T74Wk+0i/3yj6zDxPXj4r4xeZK0G218Ed78Pdt8t07wJIYTYcNVqlUwyTXpuwagKL6TJZrNk8zny5QK52upB2I0dt8VBzBHA6/Lg9XrxBfz4wgH8fQHsITcmlxXFtDnC7uXSGhrZRJmFmQLpZihOzRZJzxWpFDttIhabiUCfC9Ml6mkbQRKIuGoqp08z/8H/Re4rX8Hk8xH51f9O8Kd/GpPTufoXVQvw+D8ZrRazj4PqhVveBIffBJHd63fxQgghBFDIF5g9M8ns+Wnic3EyuYwRhOvF1YOwYsdtcdLnCuJpBmF/0I8vEsDfF8QedKE4LJum6vtsVEv1doU4NVskPWs8z8yX0Bp6+zinz0agz8muwzH8fU4CfU4CfS7cfhXFpMg8y+LaVB0fJ/HhD5P5/IOYHA7C73gHwTf+LGaPZ/UvSpw2AvIPPgWVDMQOwis+ANe9BlT3+l28EEKILalcKjN7dorZc1PMzcyRSCdIFtMU9Ur7GKtuxqs4cVud9J1Nbo4AACAASURBVHnCeFwefF4vvlCzItwfNCrCNvMG3sn60XWdQrpCarbYfLSCcYFCptNjbTIp+KIO/DEn22+ItAOxv8+J6th80XPzXbHoGbXZWRIf+d+k/+VfUMxmgm98I6H73owlEFj5C5YO2DNZYf8r4db7YPg2uIb/j1sIIcTGqFarzJ2fZvbsFHPTs8wvGKE4r3XmC7boJgImDyPOPiLBMLH+Pvp2DhLYFsWsbr2o1KhppOdb1eHu1olapdE+zmY3E+h3Mbwv2KwSuwj0OfFGHJjN184AwZ78L0Bmw+ht9WSS5P33k/r0Z9B1ncBrX0vorW/FGouu/AWlFDz6t/C9v+oM2Hvxb8FNPwPuVb5GCCGEeAbq9Trx8Vlmz0waoTiZIFFIkW0U2seYdIWA4qbfESYSCBHtj9G3fZDIzn7MDusGXv3a0hoa5UKdUq5KKV+jlKtSztco5WuUW/vyVUq5mrE/V0XvdE7gDqoE+lzs29lPoN9FIObE3+fE6bVd060lLT0ZlmU2jN7UyGRIfuKvWfi7v0Mvl/Hdey/hd7wD29Dgyl+wcBa+81H4/v+BWgG2vxDufj/svksG7AkhhHhWGo0Giak4s83llueT8yTyKdL1PHpzGQ1FV/ApTiJ2P/siO4n1GUsuR8YGsHrUDb6DK1evNoyg2wy2pUXb8pIwXMpVuwbSLaW6LDjcNhxuK76Ig74dvnZfcSDmwh9zYlW3RpvJaiSxiEuqLyyw8Dd/S+pTn0LL5/G+7G7CP/8LqDu2Lz9Y12Hiu/DtD8FTXwCTxehDfu47oO+69b94IYQQm46maRSyeVLTCVJzSeYeH+fTj/0NiVyKVC2L1lpbTgev4iSk+hgLjhCJRukbHSA2NojN79h0Vc9KqU52vkRmvkQ2WaKU7Q69pebz+qJWiMVMJgW724rDY8XuthEecuNwW7F7jDDs8NiM91vPXRZMG9Quoes62VKdeK7MfK7CfL5CPFthfrreazPHSVgWq6vNzpL8xCdI/+M/oVcqeH70Rwm/7a3Y9+5dfnCjDk/9K3z7wzD1KDgC8IJfNfqRPX3rf/FCCCF6VrlUNoLwbJL0fIpMKk02lyVbzJGrFshrpU4gbnLrdoI2H6OhAaLRKH2j/cTGhrCHXJsmFOu6TjFbNQJxokQm3gzGCWNbznfPtmGxmrB7rEbl12Ml0Odqh2GHuxl8m0HY7raiOjd+xo1KvWGE39ajFYLzxut4rkKi+V61oS37+n3B3ut1lrAslqlOTJD8q4+RfuAB0DR8r3gFobe+BXXHjuUHlzNw7O/gux+FzAQEd8LL/9RYQMTWg5MlCiGEWFP1ep307AKpmQSZeIp0KkUmmyVbyJGrFMjXi1TpbgtQdAUXqjEPsT3ELmdn1gl/OMi5+DgvevlLNsUcxFpDI5+qGEG4GYKNanGRTKLcVRVWFHAH7fgiDnYeiuCNOPA1H96wA5u9N2KaruukizXi7QBsVIOXhuD5XIVMafn0egAhl42IRyXiUdkZdhHxqkTcantf1GMn4lE59p3/u853d2m98SmInlA5fZrE/feT/bcvophM+H/ixwm9+c3YhlZYVjo9bvQjH/skVHMwejvc/T+NfuSLLWEthBBi09I0jXwyw8J00liUI5ki016dLk+uXqSoVWBJplV1Kx6zA5/NxZAnhs/rw+f34YsECfaF8A0EsThXX7jq3NHpngrK9WqDTKLUbpnoBOISuWQZTetUxc1WE96wEYCH9gSNQBx14As78ITsmC0b9ztT03SShSpz2XLzUWEuW+60RrQqwfkKtYa+7OvtVlM75I5F3DxvZ6gdgKNelYjbeC/ktmG9zHaPja6Mr0TCsqB8/DiJj/4lua9+FcVuJ/iGNxB84xtXnt1i8hGjH/n4v4JiggM/bvQjDxxa/wsXQghx1ekNncJchrnz08SnZplPJEhmF0iVs0Z7hNIdmsy6CY/JgdvqZJt3AI/bi9/vwx8K4I+FCAyEUP3OngxBq9F1nXKhRjZR7gTiVjiOF7vmFAZQnRa8YQeREQ9jN0e7KsQun7ruQb/VDzyXKzObKTcDcKUdimezFeJZIxDXte7PU1FaVWA7UY/KrpjHqP66WwG4Uw12qxvf9rEeJCxvYcVjx0h89KMUvvFNTG43obe+heDP/MzK8ySf+yYc/WO48F+g+uB5vwC3vhV8q8yEIYQQomfpDZ1GqkxuOsXc+Azz8TiJVJJkMc1CPUdR6SzMYdZN+C1uos4Qu9wefD6f0R4RDREcCOEK+zZskNiVKBdq5JJlskmjGpxNlMklS2STZbLJ8rJBdC6fDW/EwfD+oNEmEXHgizjxRRzYXes37VyxWm9XgJdWhBc/r9SX9wP7HFb6vHaiXpVd0TAxr0rMa1/0UAm71cuuAm8VPRmWZZ7ltaPrOsVvf5vER/+S4ve+hzkQIPLLv0zgp39q5RX3xr8DX/t/4fw3wdMPd70PDr1BVtkTQogepzc06qkKtUSR3EyK+PQciWSCRHaBhWqWtFKgpHQqpBbMBFUvo4EhIuEw0YEYsW0DBAcjmM2bb+qwarneDMFGAM4ljGCcTZbJJctUS9190za7GU+zXWJ4bxBPyI43bMcbNoKxdQ1X6Wv1BCcLFeZzVRJ5o/VhrlkBblWI49kKucryaeAcVjN9PiPsHhrxE/MaVeGY127s9xgB2W7dfJ9jL+jJsCzzLK+N8vHjzP7+H1D64Q+xRKNE3/WbBF77WkxO5/KDJx+Br/8RnPkauKJGSL75jWC1r/+FCyGEWJHe0KgvlKkny9QSRTKzC8zH50mkkyyUMqSVAimlQEXpDLqymiyEvH52BnYQjUWIDvcTG+nH5/Nh2kRjTmrVRjsA55rV4FyiVRkuUSl0h0qLzegd9obsDIz58YbtRiAOGb3DV3smiYamkyxUSOSqxjZvPE/kjUFxyXwnFCfz1WXtEABWs0LUY4TgPX0eXrAr0q4AGxVi4/lWaYfYKD0ZlsXVpVerJD76lyTuvx9zwE/f7/0evh9/FSbbCoMppr8PX/9jOPVlcIbgzj+EW94MthUCtRBCiDWlVRtouSqNXJVG1njUEkXS8RTzyXkWCmlSFEibCqSVAlWlExBVm42wN8i+yBDRwT6ig31EIhG8Xu+mCFaappNfKJOJl1g4rfPt1JlFLRMlSrnuWRfMFlO7Ghzd5sUbaobhZkC2u61XfN+VeqMdcpP5KvP57hDcCseJfIWFYvcqeC02i4mI2xj0FvPaOTDgJew22h9CbhsRt0rYY7wOOK/8msWVk7B8jSsfP870u99D5cQJfK/8MWLvfjdmv3/5gbOPw9H3wdNfALsfXvI7Rk+ytFsIIcRVp1Xq7fCr5arUsxWKqTz5TI5itkCxUKBQLFKuVygrNeNBlZJSJaMUqSnNfloLOG12wsEw18V2Eh2IEYlEiEQiuN3ung9ajYZGLlk2BtDFi+15h1tzD2uLZmCYM4/jDtrxhuxsvz6MpxmCvc1ZJZwe27MaSFepN0jkq+25fxP5RdtWEC4YcwNnyyuvhOeymdsBdzTk5OZtAcJulYjb1gzBKmG3jbBHxSNV4E1HwvI1yqgmf5TEX96PJRhk6CMfwfPiO5YfGH/KGLh3/F+NgXtH3gPPeTvYvet/0UIIsYnpuo5eqtPIVSkvFMgnsxTSOfLpPMV8nkKhSKlcolgtUdaqzQBco6xUqVBHXzLLBApgBYvJjNPuwOF04nVH2BENtwNxJBLB5ertOe0bNY1ssrMARyZeJDNfIt2cZk1f1H5gUc34Ig5CAy523BjGFzUG0B0/9QNecvcRTJcZhmsNzaj8Lgq/8/kVwvBFArDXbiHcnAViX5+X8JgRfFuhONwMwmG3imMN+5nFxpOwfA0qPfEkM+95D5WTJ/G98pXE3vNuzD5f90GJU0Yl+Yl/NhYPeeGvw3Pfaay8J4QQYhmt1iBxZoaZc9NkF9IU8wWKxSLFcolSrUypUaFCjRJVGsrymQgAFBTsVhWHzY7T4STqdODyuHH53Li8bpxOJ06nE5fL1X5utfb+P8W35h3OxFuhuNgMxiVyqTKLF+Oz2c34ok6iox52HY4aM0pEjYF1Tq9txXs9Na2g6TqJbGfeXyPwVpcF4ES+Qqq48sIYHrUTgPf0ebh9LGxUgJsBOOJphWEbqkUCsDBIWL6GaJUKiQ9/hOTHP44lFGLof38Ezx1LqsnJM/CN/w8e+wew2OH5vwTP+0VwhTbmooUQosfomk5+NsPMqXFmJ6aJz88zn19goZ7ttD802bBgN6s4bCoe1Uuf04nT5TQCsN+DO+jB5fO0g6/dbt9Ug+gWq5brxrLMSyrEmfkS+VSl61jVZcEXcdK308feaF+7QuyLGtOsrRSIM6Uax2eyTKVKTLYfRSZTJSaSBfJf/tKKPcBOm7kddndG3Ny2I0jEbSfs6fT/tuYGltkgxLMhYfkaUfz+95l5729RPXsW36t/gthv/AZm76JWinoVvvmn8M0/AZMFnvMOeP4vgzuycRcthBAbrJYrM3dykplzU8zNzRHPJElW0hQWzTOsYiVk93MgsotYXx/92wcJDEVw+dxYLNfGr1Fd16kUjanWcgvG1Gq5VJl86/VCedmAOofHii/iZHBPoB2ELzbvcKZU42yqyOSFha4g3HqeW9IO4bSZGQo4GAo4iZqLXL9ne3NxDFs7HIfdKi712vgMRO+S/8I2Oa1UYv4Df8HCJz+Jpb+P4Y99DPftz+8+aOoY/Os7IX4crnstvPQPwdO3MRcshBAbQKs2WDgXZ+bMBLPTs8wvJEiUUqS1fHtFOhMKAYuXoWA/sWiMvpEBBvYM4wsFer4N4lJ0TaeYrXaC8OJt83ltySIcFqsJd9CYUSI87METsuNvVYgjDmyOToRorRg3kSry/XPJZUF4KlVaNj+w02ZmOOBkKODg1m0BBpvBuBWQF88EcfToUY4c2b32f1BCrEDC8iZW+O73mPnt36Y2Pk7gp15P5L//Kmb3ooEetbIxeO9bHwR3DF7/D7Dnro27YCGEWGO6plOayzJ9apLZ8SniiTjzOaOForJoWjWXYifiDDIW2kFssI+BsWEio32btlLcqGvkU5UVQ3BuoUw+VUard/cwqE4LnpAdX8TB0J4AnpAdTzMcuwN2HJ5OWNV1nUypxmSqxOlUnskL813V4ZXCsMtmZjhohN/btge7gvBQwIFfpkUTm0RP/q0gK/hdXCNfYP7P/pTUpz6NdXiYkb/9W1y33dp90Ph3jWpy8hTc9P8Y8yU7VpgyTgghNimlojN17Cyz56eYm51jPp0gUcmQU0rtY6yYCal+dod2EOvvo3/bAP27hnG6e3sGicV0TaeUr1HMViikV64OFzKVrkF0AE6fDU/QTnTUw85DkXYQ9gSNx9LKcKpYYzJV5ESqwNSFRFd1eCpdIr8kDLtVSzP8OnjOjlD7eSsM+xwShsW1oSfDsqzgt7r8//0vZn7nt6nPzBL8mZ8h8ku/2L0CX6MOR/8HfPPPwDcMb3gAdr544y5YCCGuglqhwtQTF5g8c4HZ2VnmcgkWtBxfVx4CjFnWfGYPfYEI10di9A33M7BnmEAk1LMD6rSGRilXo5CpUMxUKWQqFDJViku2pWwVbcnqbiaTgjuo4gnaGd4bwL2oKuwJ2vEE7JitnfvWdZ1kocpkqsQTySxTp+e6K8PpEsVqdxuGR7UwFHQyEnLy3J2hriA8HHDidch8wWJr6MmwLJZrZLPMvf/9ZP75X7Dt2MHop/4e56FD3QdlJuGzb4KJ78Ch/2YsUa16NuaChRDiWaqVqkw9dZ6pU+NMz8wwl02w0Mi2e4ttWIg4gmyzeNizfy8DO4eJbuvHttKqpBugUdcoZqvtENwKvYtDcTFTpZRbeYU3u9uKy2fD5VMJDrhw+lRcPhtOr4rLr+IJqjh9atecw7quM5+vMJkq8f2ZNFPHFw+gKzKVLlGudU9n53NYGQo42BFx8YJdka7K8GCzMiyEkLC8KRS+8x2mf+M3qSeThO67j/DPvxOTqnYf9PQX4XNvB60OP/FxuO7VG3OxQgjxDFTLVaZPXGDy1Dgz09PMZRIk69n2Ah02LETtQW4KH2RwdIihfdsIDUQwmUwcPXqUW4/cvq7XW87XSM8XKaQqXZXgxZXhcn75HL+KAg6PDWczBEdHPJ0Q7FNx+VScPhtOrw2zZXklvN7QiOcqTGdKTF5ILhs8N5UuUal3h+GA08pQwMnumIc79kQ7leGgg0G/A49dwrAQl0PCco/LfuUrTP3qr2EbGWHbhz+M47qD3QfUK/DV34HvfhT6b4BX/zWEdm7MxQohxEVUq1WmT44zefIC01PTzGXmWajl2sFY1a1E7QEOxw4wMNIMxsPRdW+jqJbqpJtLL6fjxa7nlUJ3367JpLRDrifkoG+nv1kFtrUDsMuv4nBbMZlXvg9N00nkK5zJlJmdTjGdLjOTKTGdKTOTLjGTKTOXLbOkE4OQy8ZQwMG+fi937o81Z5NoVob9DplSTYirRH6Selj6c59j5j3vxXH99Qz/5UeXr8KXPAOffSPM/BBuezvc+ftgUVc+mRBCrKNKpcL02UkmT1xgemqKufQ8qVq2PQbNrluJqgF29o/QPzzI8N5tBLfFVg2UV1ut2miuNldsBuLW8xKlbLXrWHdAxR9zMnZzDH/UgS/qNFohvEYIVi6yBHNr4Ny52RwzmWYIbobhmXSZ6UyJuWyZWmPJTBUWEwN+B/0+O8/bGWbAb6ff56Dfb2c44GDA78Bpk1/h4tqjr9SbtMHkJ61HLXzqU8z9wR/ifM5zGP7whzC5lozcfuyf4Au/bCww8rpPw96XbcyFCiG2vEqlwvSFSSZOnGdmcprZBSMYtzh0GxGrn7HICAPDgwztGSW4M4ZpjZcTbtS05hLMRgg2KsRGlXjpinNOrw1/zMm260L4o05jPuHmEswW28rXqes62XLdCL6ZMjPpJWG4GY6X9gpbzQoxr50Bn4ObRwP0+xydMOyzM+B3dM0xLMS1pqbVmMpNMZ4b50L2AheyFxjPjjOeGyfUCHEHd1z6JOtIwnIPStz/V8z/2Z/hvuMOBj/w5939ydUCfOk34Pv/B4afA6/+OPiGNu5ihRBbSq1WY/rcJBMnzzM1OcnsQpxUNdd+36mrRCw+doVGGBgcYGjvKMGxPkyrBM4rpTU0sslyu00isygU55LlrgF0dpcVX9TB4J5Au0LcCsU2+/Jfh632iMnZ7LIV56bTJWbSJQpLZpAwKRD12On329nf7+Ule6P0+x0M+OztbdjdPThPiGtRXaszU5hhPGsE4vHcOOez5xnPjjOdn6ahd352PFYPI94Rro9cjzvj3sCrXpmE5R6i6zrzf/4Bkvffj/flL2fgfX+MYl00AGPuOPzTz0LiJLzg1+DIu8EsH6EQYm00anVmT00yeeoCk1NTzKbiJGoZ9GYzhVNXiZh87PaP0D84wNDuUYK7YpicV3/gmK7pZOZLJKfyJKbyJCfzTJ3VeOofH+qaVs1mN+OLOolt87L7tr52GPZHncuWYNY0YwaJx+eyiwbMdQbNTaZLVFcYNDcYcLAj7OL2sTD9i0Jwv99BzKNiWadWEiE2mqZrzBZm25XhC7kL7XA8mZ+krnV6/J0WJ6PeUfaH9nPXtrsY9Y4y6h1lxDtCQA10rdbYayRp9Qhd05j7o/9B6u//Hv9rXkPf7/0uirlZidF1OPa38KXfBNXbnDu5t/6JQgixudXzVZKnZ5k4fYHpmWlmM3HitTR1xaj+2HQLEaufm0P7GBgcYHin0Uphdl/96doqxZoRiicLJKfy7Ue9agRXRQF/zIndB2PPG8Ef61SJF686p2k68VyFyVSR75yMd80p3Fp1rtroDsOtQXN7+z38yP5Y93RqMmhObEGarhEvxpeF4fHsOBO5Capap8ffbrYz7B1mzD/Gi0dezDbvNka8I4x6RwnZQ5u2tUh+6nuAXq8z81u/TeZznyP4xjcS/Y1f7/wHVSvDg78Ej30GdtwBP34/uKMbe8FCiE1Lb+jUE0WyF5JMnh1namaa2ew8cS1NWTGmPDOhELEFOBDbxeDQICO7txEdG7jqPcaappOJF0lMNgPxpFE1zi90+olVl4XwkJv9tw8QGnQTHnIT7HdhsZn52te/zuihASZTJZ5OL1qCOV1st0osHTgXdtsYDDjZP+DlpQdixlRqfiMQDwZk0JzYOqqNKtlqlkwl07XNVrLMl+bb4XgiO0G5UW5/nc1kY9gzzIh3hBcMvcAIwx6jQhx1RjEp196/rMjfChtMq1aZ/rVfJ/eVrxD+hZ8n/I53dIJyPg6f+WmY/B7c8V6j9aJHV6ISQvQerVijOlOgOJlh+twE0/EZZvMJ5smSM3WWhA7avOwMbmNweIjh3dvo3z6IxXJ1fz2U861qcadSnJwu0GgOflNMCoE+J/07/YRf5CY0aDzsXiszmTLnEgUeThQ498MJzn+twPlEgYmFIo0vf63r+0Q8KkMBB9cP+XnZdf0M+rsrw4416p0WYiPUtTq5aq4dcjPVTNd2xTDcPHZxAF7KYrIw5B5i1DvKc/qf0w7Do95RYs4YZtPW+jmSsLyBtFKJyV/8JQrf/CbRd/0moZ/92c6bs0/Ap18HhQS89pOw/5Ubdp1CiN6mN3TqyRK1mQKV6Rxz49NMz88yV15g3pQlpRTacxl77C76Q/0cHhliZNc2+ocGsNvtV+1atIZGaq7YrBQX2uG4kO5Uix0eK6FBNwdfNEh40E1gwEXVZWYiXeJsosD3EgXOPxrn3JcLTCx0t0o4bWa2h10cGPRxwFvj+TfubbdKDPgd2K1b65e42LzqWp1ivUixVqRYL1KqlSjUCu19hXphWeBtvW7ty9fyF/0eDosDr82LV/Xis/kY8YzgU314bd72tvXe4q3b6t5ygfhiJCxvkEY+z8Tb3kbp0WP0/eEfEHjNazpvPv1F+Oc3g90HP/fvMHDjxl2oEKLnaJU61Qs5yufSzJ+ZYXp2mnktw7wpS0LJ0VCMcGm3q/RH+jgwegND20cYGBjA4/FctesoF2rMT+RITnZaKBZmCmh1I5ibzAqBPheDe/yEBtxYQipZG0yUKhxfKHI+keH8d2Y4nyx0rT6nWkxsC7nYFfVw5/4+toedbAu52B52EfGoXQOBjtw2ctXuR4jVrBRsi/WiEW6b+y62LdSN40r1Unt/pVG59DfGqPIuDrMRZ4Qx/xhe1dsVetvhd1EAtppllcaroSfDsqIo9wD3jI2NbfSlrIl6KsXEfW+h/PTTDP7pn+B9WXOOZF2Hb30Qvvq7RkB+3afB27+xFyuE2HCNXJXyuTQLJ2eZPD/BbDrOvJIjYcpSVepgAovFTF8kxuHRvQwNDzEwMEAwGLxqA2o0TWdhusDcuQyz57LMnc2Qmi2233d6bYQGXey+fYCa20zKCpNajScWSpxLLHDh3ETXNGs2s4nhoIPtYTcv3B1mW9jF9pCLbWEXfV67TK0mNkSmkuHJ5JM8mXiSJxJPcHzhOKly6rKDLYBqVnFanDitThwWB06rE5fFRcQRae93Wpw4rA5cFlf7dWvrsrpwWB04LU68Ni8Oi2PTDoy7VvRkWNZ1/UHgwcOHD9+30ddytWnVKuNvehPV02cY+l8fxHNHc1aLegW+8Cvwg7+HA6+CV34EbM6NvVghxLrTdZ16okTqxByTJ88zPTPNXHmBhClHSTFGnZusChF/mIOj1zHYDMbRaBSz+er9s2kxW2XuvBGKZ89liZ/PUqs0Z8ZwWXD2O3EfDrFgVxhv1DiTK3EuESc305kqymxSGA442B52cduOINvDrnaFeMDvwCyBWGygcr3M0wtP80TiCR5PPM6TySe5kL3Qfn+bdxs3RW8i5owtC7ROqxFqW6F3cTi2mqSae63pybB8LUt85CNUjj/F0Ic/1AnKhQT8wxtg/FvwonfBkXcZcyMJIa55ekMndz7B+BPnmLowwUwqTkLLkFc6g29CXj9j/WMMjY0wODRELBbDar16v5AbDY3kZJ7Zs1nmzmWYOZshl2h+fwXqXgvpgIlxRed4pcxMowSJHCSMv6oG/UYgftWhwXYY3hZ2MRRwYJU5h0UPqGt1zqTP8HjicZ5IPMGTySc5lTrVXhgj6oxyMHSQe8fu5UDoAAfCB/DavBt81aJXSFheR6UnnyT5Vx/Dd++9eF7yEmNn/Cn41E9Cfg5+4uNw3as39iKFEGuqnC8x8dhZJk9dYHp2hrlikqzSaWfwWd0MhQcZ3D5szEzR34+6eBXPqyCfqnDu5AJnn15g/nyWSrwMzSnWimaYNDWYtmtMWzTmzBpmq4kRr5PRkIu7g1FGQ05Ggk5GQk6GAg7UNV62WohnQtd1xnPjPJF4ov14euHp9uwPXpuXg+GD/NzBn+Ng+CAHwweJOmVKVrE6CcvrRK9WmXn3e7AEg8Te/S5j56mvwj+90Wi3+NkvwtDNG3uRQoirql6vM3Nukonj55gcn2Q2PU+qnkVv/sORS7HT549ww+B1jOzdzuDYCE7n1Wm/ai3IcW4ux9lTKebPZSnNlbCla9hrzetDZ86sMWPRyXpNWKN2Yn1GKL49aGxHQ04isjyz6GHxYtxoo2j2GT+RfIJccwl2u9nOvtC+/5+9O4+Pqrz7//86M5NlMpmZZDKZyb6QELITSAKyKSIIKq5FBbWW0n6rfmttXW7rcvdXu92PavfWbtZWv93Qu4obKgpiZBHZQyAJSSAkZN8myeyTWc7vj4ERFJAlG+R6Ph55JDM5c85nMiF5c+VzXRe3TrmVwrhgME7VpooeYOGciLA8Snr/9Gc89fWk/OEPKHU62PYHeP9JMBfCipdAnzzWJQqCcAH8fj89PT20NjTTcqiZju5Oel39BI5tDR0ph2GKNDA5MZOUSamkF2ejM8Zc0DXdXj+t/U6OWpw09zk52uegs92Bq9NJ2IAXk1eB2S+hRCIMGFKBXafEPZxhvAAAIABJREFUHR+JIU3L5Cw9C03RpBmi0EaKPkth/DtxAt7xgNzt6gZAKSnJic1hccbiUDDOislCpRBRR7gw4jtoFLhra+l97jl0N1yP9sr58O6jsOM5yLsebv4zhGvGukRBEM5RIBCgq6uLw/vrOFx/mNa+DrxycHJbmKwkXtJTosshKSWZtIJJGHMSUZzjGsCBgEyP3cNRi5MWSzAUn/jxwIAHs19Bgl9Bkk9BckBBYSA4YiYrw1AZIzCkaUmbHENeoRG9QT3sXwdBGC4BOYBtyIbFbcHittDv7g+9/6TnE37+2s8/NwGvPLGcImMRBXEF5BpyiVQN35rhgnCcCMsjTPZ6aX/iSZQxMSQ88URwtYsdz8Fl34Srfyx25BOEi4Qsy/T19nGoqo7G+sMc7WnDHQguJ6UPRDFZlURKQhIp2Wkk5KcRZtYgnUXrgsPjo6XfSYvFdcpQfHz94TAZzAEFk1Xh5KFijltFmOvTnx/a+EiSJsWQMEmHOVNPXLIGhZhcJ4yhgBzA6rFi8ViwuCz0e/pDAfh4CO5394c+P+AZCE24+6wYZQylxlIxAU8YEyIsj7Dev/wFT20tKc/+DqW3G975L8i8HK7+kQjKgjDODQ4McqjyII11h2jubsXuD24RrZEjSFMaSU9KJSs3G2NBMkpD5Cn7IP0BmU6rOxSCPxuGe+1DJx0fHaEiPUZNYWQkixM16F0BFP1ePBYPxzo60MSoMOVoMWXoMKVrMaXriNSINgphZAXkAIOeQfrd/fS5+z4Xdj8bhgc9g6cNv9pwLYZIA4ZIA6nRqRQbi0O3YyNjiY2MJS4yLvhxRCxbN29l/vz5o/uEBeEYEZZHkLuujt4//gndddehvWIuPL8QVJFw83MgtpEUhHHHYXdwaO9BGg8Gw/GANzhJKEIOI1llpDyxkKzcbMyFaYTFfdrSMOQL0NBpo7nPEQrCRy0uWixOWvudeI+tNAGgkCApRk2aIYqFeWZSY9QkSEqiHQGwDDHY7qDvkJ2A3wGAMjoMU7oO08zEUDjW6Id3dQzh4tBh76C6rxqXz4Uv4MMb8OIL+IJvsg+v34tPDt4+8ePjx554/Ok+Pu3nZB8ev4eAHDhlbdpwbSjcpmnTKDGVEBsRGwq/JwXhiFixs5xwURFheYTIXi8djz+BUqfD/N9PwoanoGs/rHhZ7MonCOOEx+WhsbKOQzUNNHe10Ds0AAR7jhNVcRQmZJM1JZuUkkmExaqR5eAo8ZYOG7X726jrtHGww8bhHju+wKeBWK8OI80QRX6ijsUFCcFl1gxRpMREovGCpcVOT7ON7kYrPS0WOoeCASQ8Ukl8uo6pV6UGA3KGFu1pRqyFS5ssyzRZm9jdtZs9XXvY3bWbdkf7WT1WpVARpghDJak+/VihCr199rZapf7c50LHSCrClMFzRagigoE3IvakABwTGSM24hAuaSIsj5C+v/4Nd00Nyb/5Daru7bD9jzDzPpiyZKxLE4QJy+vx0rSvgcPVDTR3HqXTbUGWZBSyRILKwCxzMZNyskmblo1PE05dl41tHTbqNh2mttPGwQ4rVvenO9Ql6SPJTdRxVZ6JKQlasuKjSY2NQh8VhizL2CzuYChuttL1cR+1R20MuYKPV4UpMKZqyZ+bFAzG6VpiTFFn1ecsXHr8AT8NAw3s7toderO4LQAYIg2Umku5u+BuSuJL0IZrTxt8VZJK/OdKEIaZCMsjwNPQQO/vf4/2miXoZhXBn+ZAQhEs+sFYlyYIE4rf6+NoVSOHD9TT3HGUdlcvfimAJEO8MoZScz6ZWZMIm5TGIbeP6k4br7ZZObh7O0ctn24UoglXMiVBy9KpSeQmaMlN0DElQYteHRxNk2UZp3WInmYbB/e00N1ko+eoFZctuKCxQikRlxzN5HJzqMfYkBglJuBNYF6/l+q+6lAwruyuxHas7Sc5Opk5SXMoNZdSai4lXZcuArAgjCERloeZ7PPR/sSTKKKjSXjyCVjzVfC64Et/A5XoMxSEkRTw+fF2OPjon+toamumzdnDkBQcyTUotOTHZRNlSqY7Jp5qu5d1nTbqt3Tj/rATCPYTZxg1FCXrWVaaQm6ClrxEHckx6tCmHE7rEJYOB82fdGHpcGBpt2Npd+BxBq8jSRCbqCG9yIgpLTgJz5gcjTJMBOOJzOl1UtVbFWqpqOqpCu0oN0k/iSWZS5hunk6pqZTEaNGqJwjjiQjLw6zvhRdw799P8q9+iar279C0GW54FuJzxro0QbgkuS0O6rbtp76unqbBNhxScDk3rRRFQnQynqh49kta3un30t3ugXY7YMegCScvUcsdM9LJTdSSl6BjsjmayGNrIbvtXiwddiz7LRxpd2DpcNDX7sBt94auHRGlwpCkIbvUhCFJgzFVS3yqlrAIMYF3ohv0DFLZXRkcOe7eTU1vDT7Zh0JSMCV2CstyllFqLmWaaRpx6rixLlcQhDMQYXkYeQ4fpve3v0O7aBHaAgO88BMo/BJMu2usSxOES4YckOk92E7trv0cammkbSjYWhEmK4lWGenxR7PVo6VNDgMXhA8omGxWMG9yfLCFIlHLlAQt8dERSJKEx+nF0u7AcsTOzq3HR4sdOK2fLukWFqkkLknDpKlGDEnRGBI1GJI0ROnDxZ/HBQB6Xb2hloo9XXuo769HRkalUFFkLGJl4UpKzaWUxJcQHR491uUKgnAORFgeJrLfT/sTT6CIiiLhu99B+s+NwS2sl/4q+HdZQRDOm885xOHtB6neV0PLQCv92AGIlNX4/GYOBPTsDWhIiIwiPsrLTbMyyE3QkZeoJSNOg0qpYMjtCwXh+soWPjn2sWPAE7qOKkKJISGKtAJDMBQnaTAkaoiOjRChWAiRZZl2R/tJk/GO7yynVqmZGj+V/1vyfyk1l1JkLBK7ygnCRU6E5WFiefH/4d5XRdLPf4Zq24/A2gar3oNI/ViXJggXHdeQj4P72zm0s5r+7qP0+XvxSF4kWSIioMMmZ9CuTyAhLYGCJB2LE3XkJeqI1YSzccOHFE5OwtLuoGtzF7XtDiwdduyWE0JxmILYRA0pU2JDgdiQpAku0yZWo5jwHF4HXc4uuhxddDu76XZ2B287g7c7HZ2hlSp04Tqmm6azbHKwrSI3LlcsoyYIlxgRloeBp/EIPb/9LdFXXYUu0QJvrYGr/j9ILR/r0gRh3Ot3DFHbYaWmZYDB2haUPW1I/l4GpEFkSSZMDiNCaSQiPpmEohyKMswn9Ra77EO01w+wf+cRWusH6O+QqWUXAAqVRGyChsSsGAzzPg3FOuOnE/aEiSMgB+h394dCb5fj0wB8/H23sxu71/65x+oj9JijzJiiTOQZ8phimEKpuZTsmGwUkpi8KQiXMhGWL5Ds99Px5JNIkZEk3H8n0ppbgttZz/nOWJcmCOOKLMu09ruobrdS02Glpt1KZ0s/k539pCsG8Cks2BXB1QG0Si2TjHlMKS1gelkuKtWnE+bcDi/t1Rba6vppq++nry24050qQklSlh5VnJPSOQUYEjXo49ViebYJwuv30u3qPm0I7nJ00e3qxhfwnfQ4paTEqDZijjKTFZPFrKRZmKJMoWCcEJVAfFS8aKUQhAls1MKyJEka4A/AEFAhy/K/RuvaI8mxdSuuvXtJ/OFThFX8F4SpxXbWwoQ35AvQ0G2j5oRgXNNhxeH2UYSfMqWVacpBBiQL3nA/NhQkRycwe0oOeZcVExtvCJ3L4/LRWtNLa30/bXX99LbaQQZlmILELD0zbzCTPCUWU4YWpVJBRUUFWdNMY/jshZEQkAN0OjppsjbRNNhEs7WZqu4q/vDWH+hydoXaIk6kVqkxRZkwRZmYbp4eCsHHg7BZYyYuMg6l+HktCMIZXFBYliTpb8BSoFuW5cIT7l8C/AZQAs/LsvxT4BbgFVmW35Ik6WXgkgjL1nfXodBq0al3BrezvuN/xXbWwoTi9Qeo77Kxv3WQqrZBqloHqOu04fUHt382qhRcHy2zONyCQ+6mT7LiB5wKNXkJk8mdmk/2tFzCw8MBGHL7aD7QFxo57jlqQ5ZBqVKQMEnHjKWZJOfEYs7QibWLL0GDnkGarc2hUNxkDQbjo9ajoXWJAaJUUcRIMWSps8iPy8esOSEEH3uvC9eJiZmCIFywCx1ZfhF4Fvj78TskSVICvwcWAa3ATkmS3gRSgP3HDvNf4HXHBXloCNuGDWjLc1Hs/ktwO+ucxWNdliCMGH9A5nCPnarWQfa3DlDVNkhNuxWPLwCANlJFcbKO7xQnkWrtxt7bRqu7A6d7iKNAQlQc8zJmkjeziMT0ZCRJwjvkp/PQYGjkuLvZhhyQUSglzJk6Sq/NIDknloRJOlRhYgTwUjDkH6LF1nLSKPHxUHziCLFSUpKiTSFDl8FliZeRoc8gQxd8M6qNfPTRR8yfP3/snoggjBNer5fW1lbcbvcXHzzO6fV6amtrR+z8kZGRpKSkEBZ29hNxJVmWL+iikiRlAGuPjyxLkjQLeEqW5cXHbj9+7NBWoF+W5bWSJL0ky/Ly05zvG8A3AMxmc+lLL710QfWdid1uJzr6/Ne7DN+/n9jf/4HEK50os2LZM/1nyGIW9Lh0oa/1RBSQZbqdMkcGAzQN+jliDdBsDeA59l/dSCWk6xRk6hVk6JTkBgIoeix0DfbQIVsISDJhKDFFxmGIN6BNMRAeEU7AJ+PsA0eXjLMbXBaQA4AEagNozKAxSUQZgxP0zpV4rccHWZYZ8A/Q7e2m29cdfH/s4z5fHzKf/u7RKrSYw8zEh8VjVpkxhZkwhZmIU8Whkk4/piNe64lDvNZnFh0djdlsRq/XX/R/TfH7/SiVIzMwIssyg4ODdHV1YbefPJH3yiuv3C3LctmpHjcSPcvJQMsJt1uBmcBvgWclSboOeOt0D5Zl+TngOYCysjJ5JEcNKioqLmhUov3dddiiItAb25HufJMrEqcOX3HCsLrQ1/pSd3zy3f62Qapag60U+9sGsbmDk6EiVAoKknSsyIuhKFnP1FQ9GQYN/bVtVG+vor6hjn0BC7IEWmUU01IKySsrIjM/CwISXU2DtNYN0Hagn64jVvy+AJIE8ek6pkyPIXlKLIlZesIjL/xHknitR5d9yB4cIT7FKLHL5wodp1apSdelU6YrI12XHholTtelow3Xnte1xWs9cYjX+sxqa2tJSUm56IMygM1mQ6s9v58JZ0Or1WK32ykrO2UuPqVRm+Any7ID+OpoXW+kBYaGsG3ciDZTgZRUBCIoCxeRzkF3KBBXtQ6yv20QiyO4Y12YUiIvUccNU5MoTtFTlBxDjjkalVKB7AvQvvcINWs+4L3OI/QwCIAhTMdlGaUUzi7BnJpET5ONtvp+3tpQRWfjIH5vACSIT9VSND+Z5CmxJGXHEK4WC/JcTGxDNmr6aqjuq6a6t5rqvmra7G2hzyskBcnRycFQbC4Ltkzog4HYHGW+JH6RC8J4Jf59nZ3z+TqNxG+qNiD1hNspx+67pDi2biVgs6GL64PiJ8e6HEE4rV67Jzj5rnWQ/W0DVLUO0m0LbtChVEhMNkWzKM9MUYqe4hQ9UxK0RJywVJvf46NlWx01ew/Q0NtEvxRcqs0UaeDy7MsonD2NiDAdLTV97FtnobWuEa872KsRlxJN4bxkknJiSJocQ6RGtCldLJxeJ7WW2lAorumrocnaFPp8cnQyBXEFLMtZRqY+kwxdBqnaVMKV4WNXtCAIwggYibC8E5gsSVImwZC8HLjjXE4gSdL1wPXZ2dkjUN7wsK1bhyIqHI15CAqXjXU5ggAEl2w70D7I7qZ+9hztp6p1kLaB4J/CJQmy4qOZm20MjhinxJCfqEMd/vneML/TS+O2Gmqqqjk00IJNciEBSRoT03NLmVJahKMXjtZYeP+PjQz2BK+hNUQyudxMWp6B5JxYIqNFOL4YePwe6ix1HOg9EArGjYONBOTgxE1zlJmCuAJuyLqBgrgC8uPyiYmMGeOqBUEYTyRJ4s477+Sf//wnAD6fj8TERGbOnMnatWt58cUX2bVrF88+++xJj8vIyECr1Yb6lC+77DL+9Kc/jXr9Z3KhS8etBuYDRkmSWoHvy7L8V0mS7gfeI7h03N9kWa4+l/PKsvwW8FZZWdn/uZD6RkrA48H2wUa06X6kSXNAnzzWJQkT1IBziN3N/exq7md3Uz/7WgdCK1OkGtRMT49l5ewMilL0FCbriY44/T/5oUEX9VsOcLC2lsO2VlzSEAok0vSJzM2bjTkjm75mD0erLRx47wABv4wqXEHylFiKF6SQlh+H3qQWfwoc57x+Lw0DDSe1UhzqP4RPDvanGyINFBoLWZS+iEJjIflx+RjVxjGuWhCE8U6j0XDgwAFcLhdqtZr169eTnHx2+ejDDz/EaAz+nLHZbCNZ5nm5oLAsy/KK09z/DvDOhZx7PHNs3UrAbkdn6oPiR8a6HGGCkGWZpj4nu5osoYB8qDs4m1elkChI1nPXZemUpcdSmh6LSffFO465eu0c3LyPg/V1NDnb8Ug+VCjJNKSQnTMFtSaJzgYH1Rst7LIFl/IxpkZTsjCV1Pw4EifpxVrH45gv4KNxsDEUiqt7q6nrr8Mb8ALBLZwL4gr4auFXKYgroMBYIHqLBeEi94O3qqlptw7rOfOTdHz/+oIvPO7aa6/l7bffZtmyZaxevZoVK1awefPmYa1lLIjZNefBum4diqgwNEky5N841uUIlyiPz8+BNiu7my3sOtZW0WsPTsLTRaooTY/l5mnJlKbHMjUl5pTtFKdiaxugdkslBw/Xc9TTiU8KECGFMSk+neTkLGRvLG11Nna9YQcaUWvDSM03kJZnICXPgEYfMYLPWjhfATlAk7WJ6t5gG8WB3gMctBwMbeQRHRZNflw+d+XdRb4xn4K4AlKiL43Z84IgjA/Lly/nhz/8IUuXLqWqqopVq1adVVi+8sorQ20Yt99+O48//vgXPGJ0jcuwPJ57lgMeD/YPNqJN8SBNuRrUsWNdknCJsDiG2HO8paLZwr7WQYaOtVSkx0VxeU48ZekGyjJiyY6PRqE4u5AjyzIDR3qo3rqPuuYG2rw9BCSZKCmCKaZsYvVpOC0aOuoG6T3gRqHsJDFbz2U3TSItPw5jSjTSWV5LGHkBOUCHo4OmwSaODB7hyOARGgcbqbXU4vAGJ1+qVWryDHksy1lGgbGAgrgC0nXpKCTxVwBBuNSdzQjwSCkuLqapqYnVq1dz7bXXnvXjLuk2jJEynnuWHVu2EHA40CX2Q5GY2CecH1mWaex1sLupn13NFnY199PYEww6YUqJgiQ9d1+WTllGLNPTYzFpv7il4rPn7zvYzv5t+6hvPUSn/9gayIoocuOmoFYl0Neqonv/EN0MEWNWkTcnibR8A0k5McOy3rFwYZxeJ83W5mAgth4JheNma/NJ2z5rw7Vk6jNZOmkphcZCCuIKyNRnolKI11AQhNF3ww038Mgjj1BRUUFfX99YlzMsxE/Tc2R9dx3KqDA0qeGQs2SsyxEuEm6vnwNtg+xq7g+1VBxf1zgmKozStFi+ND2FsvRYpqbGEHme2zr73F4OvL+L3VV7aPH1ABCr1JKryyPgjsPSKtHTLhEeKZGSqyd1iYG0fAM6o3rYnqtw9mRZptvZzRFrcIQ4NFpsPUKnozN03PH1izN0GcxMnBlaqi1Tn4kh0iBaKQRBGDdWrVpFTEwMRUVFVFRUjHU5w0KE5XMQcLuxb9yILsmBVHgDhImAIZzagHOIXU397DzWb7y/dZAhf7ClItOoYUGuibL0WMoyYplkPPuWitPpb+5h53tbqWo/iB03UUSQHZaDq9/AkD2cXglM6TrKrgmGY1OmDqVS/El+tHj8ntAocdNg00nh2Olzho7ThGnI0GVQZi47KRCn6dKIUIpecUEQxr+UlBQeeOCBU37uxRdf5PXXXw/d/uSTT4CTe5bz8vJYvXr1yBd6DsZlWB6vPcuOLVsIOJ1oU2xQdOtYlyOMI+0DLnY2WYJvR/qp6wr2XIUpJYqS9ayck0HpsVUqjNHDE3oC/gCHthxg5/YdHHa0EZBk4olDb5uEx2HAq4skqziO1HwDqbkGsebxCJNlmT53X6iPuMn6aU9xu70dGTl0bKImkUx9JjdPvjkUiDP1mcSr48UosSAIFyW73f65++bPnx/apnzlypWsXLnyc8c0NTWddFv0LJ+l8dqzbH13HUq1Es2kGMi8fKzLEcaILMsc7rGz40g/O5ss7DhiCW38oQlXMj09lqXFiZRnGii5gJaK03Fa7Ox6dyt7D+2nX7YTJqtI8CTjtycQpY0la6aJrOnxJEzSi4l5I0iWZY4MHmFH5w52dO5gd9duLG5L6PORykgy9BkUG4u5MetGMvTHRom1aUSFRY1h5YIgCMK5GJdheTwKuN3YNm5En2xHKl4BiuENQML45fUHqG63svNIcOR4V/On/cbG6HDKMwx8bW4mMzIN5CZoUY1Qe0Prvka2bdxK/UATXsmP3h9NiiuH8IgUssuTyJpuwpShFSOTI0SWZVpsLaFwvLNzJ72uXgASNAnMTZ5Lflw+mbrgKLFZYxarTwiCIFwCRFg+S/ZNm5BdLnQpDtGCcYlzDvmoPDrAjmNtFXuPDuAc8gPBJdwW5Jooz4ilPMNAplEzouHU6xpi11vb2FtbSbfcj1JWYPKaiFamM2n6ZLJLzRhTo0VAHiEd9g62d25nZ+dOdnTuCE26i4uMY0biDGYkBN9StaniNRAEQbhEibB8lmzr1qFUK4jKS4WkaWNdjjCM+h1DoX7jHU39VLcN4gvISBLkJui4tTSF8kwD5RkGzGexK95wOFp1lE/WbaHRcQS35CU6oCaLHCYVTSV3djqGpJEN6RNVj7MnNHK8o2MHrfZWAGIiYihPKOdrhV9jRsIMMvWZ4usvCIIwQYiwfBYCLhe2Dz9En2RDmno3iF+SF7XWfie7mvqDI8dHLDQc2zI6XKlgaqqeb1w+ifJMA9PTYtGrR29SnKXTzu63dnC4pYZOqRcJMMvxZCXnM31pGXFJ2lGrZaKwuC3s7NzJzs6dbO/YTpO1CQBtmJbShFLuzLuT8oRyJsdOFi0VgiAIE9S4DMvjbTUM+6bNyC43ujSX2IjkItTc52DjUS+vvbSXnUcstA8GN3TQRqgozYjlpmnJlGcYKE7RD/tkvC9i6XBwcFsThysr6aQFh8JNpBROQfQUZiyeQ3px2qjWc6kb9Ayyq2tXqK2iob8BgChVFNPN07ll8i3MSJxBbmwuSjEvQRAEQWCchuXxthqGdd27KNUSUSWFEJc11uUIZ+FQt41393fy7oFOajqsAMRr+5iRYeAbGbGUZxrITdChHOXVImRZxtLuoGF3F42767F6m+gO6yGgCGCWDMwtmEX59bNRRYhl3oaDfcjOnu497OgItlYctBxERiZSGUmJqYRrpl3DjMQZ5MflE6YQX3NBEIQL0dXVxYMPPsgnn3xCbGws4eHhPProo9x8883ndJ6mpiYyMzN58skn+fGPfwxAb28viYmJ3HPPPTz77LM89dRTREdH88gjj5z0WKVSSVFRUej28uXLeeyxxy7oeY3LsDyeBJxO7B9+iD7FjjT11ItsC2NPlmVqO2ysO9DBOwc6OXSstaI0PZb/vi6PaGsTt1975Zj0mcqyTG+LnUN7ujm0u40hVytWdTuDSjthYUryY7KYtXAOyUUZo17bpcbpdXLQdZDK3ZXs7NxJdV81ftlPmCKMqfFTuW/qfZQnlFMcX0y4MnysyxUEQbhkyLLMTTfdxFe+8hX+/e9/A9Dc3Mybb755XufLzMzk7bffDoXl//znPxQUFHzh49RqNZWVled1zdMRYfkL2DdtQnZ70KV5oPCWsS5HOIEsy1S1DvLOgQ7WHeikuc+JQoIZmQa+fFkBiwsSSNAHJ+RVVBwd1aDsdnjpODRAW90Ajft6cA5aUGg76AnvxBvtxyBpWZgzh9Ils1DHRo9aXZcSWZZptbVS2VPJvp59VPVUUd9fj1/2o+pRUWgsZFXhKmYkzqAkvoRI1ehMzhQEQRhT7z4GnfuH95wJRXDNT894yMaNGwkPD+fee+8N3Zeens63vvUt/H4/jz32GBUVFXg8Hr75zW9yzz33IMsyjz76KO+++y6SJPHf//3fXHvttQBERUWRl5fHrl27KCsr4+WXX+a2226jvb19eJ/bWRBh+QtY161DqYaombMh2jTW5Ux4gYDMnqP9vLO/k/eqO2kbcKFSSMzKiuPeK7JYlG8eth3yzoXb4aW9YYD2+gHaGvrpbbUjyzKRGgsebQc9RgsKWSI7OpXyWTPImpWPQmw3fU5cPhcHeg+wr2dfKBwf3wQkShVFkbGIVYWrUHWpWLlwpdj4QxAEYRRVV1czffr0U37ur3/9K3q9np07d+LxeJgzZw5XX301e/bsobKykn379tHb20t5eTnTp08nLCzYFrd8+XJeeuklzGYzSqWSpKSkLwzLLpeLkpKS0O3HH3+c22+//YKemwjLZ3C8BSMm1YE09baxLmfC8vkD7Dhi4d0DwYDcbfMQrlRweY6RBxflsDDPREzU6P5J3W0PhuO2+n7aGgboa7ODDMowBaaMaFILnBzuq6VXdqBFzZyU6cxYMgd9Styo1nmxkmWZVntrMBh3B8Px8VFjgHRdOnOT5zI1fipT46eSHZMdmpBXUVEhgrIgCBPXF4wAj5ZvfvObbNmyhfDwcNLT06mqquKVV14BYHBwkIaGBrZs2cKKFStQKpWYzWauuOIK9uzZw8yZMwFYsmQJ3/ve9zCbzWcdeCdMG8Z4WQ3D/tFHyJ4htJkByFs6prVMNM4hH5vqe3i/uosPDnYz6PISGabgyikmlhQmsCDXhDZy9CZkuWxDwXDcMEB7fT99bQ4AVGEKErL0zLw+E2O6hqYDlew8uAGHzY1R0nF9yUJKrpmJUkzYOyO3z011X/VJ4bjP3QcjyLGuAAAgAElEQVSAWqUOjRpPjZ9KcXwxsZGxY1yxIAiCcKKCggJeffXV0O3f//739Pb2UlZWRlpaGr/73e9YvHjxSY959913z3jO8PBwSktL+cUvfkFNTc159z9fqHEZlsfLahjWd95BqZaJmrsQIsQatyOt1+7hg9ou3q/uYvOhXoZ8AWKiwrgqz8TV+WYuz4knKnx0vmWd1qFjbRXBkWNL+7FwHK4gMUtPdpmZ5MkxmDJ0uB1Otr75IRt2VOHGS6IyjmtnLCTvqukoVKLV4rNkWabd0R4Kxft69lFnqcMn+wBI06YxO2l2cNTYFBw1VinG5Y8qQRAE4ZgFCxbwxBNP8Mc//pH77rsPAKfTCcDixYv54x//yIIFCwgLC6O+vp7k5GTmzZvHn//8Z77yla9gsVjYtGkT3//+908678MPP8wVV1yBwWAY9ed0nPgNdBoBhwP7Rx8Rk+5EKrmwXhfh9I70Olhf08n71V3sPtqPLENyjJo7Z6ZxdX4C5RmxqEaht9dpHaKtvv9Ya8UA/R3HwnGEksQsPTkzzCTnxBKfpkV5LAAPdlt454U1VLXV4sVPekQCc+fNI3tOvtjd7QRun5uavppQMN7Xs49eVy8QHDUuiCvgKwVfCY0ax6lFq4ogCMLFRpIkXn/9dR588EGeeeYZ4uPj0Wg0PP3009x66600NTUxffp0ZFkmPj6e119/nZtvvplt27YxdepUJEnimWeewWw209fXFzpvQUHBaVfB+PGPf8yvf/3r0O3W1tbP9SwvWbKEn/70wlpTJFmWL+gEI6msrEzetWvXiJ2/oqKC+fPnn/Jz1nfeoe2hh0m/xkPUzw6CSiwzNRwCAZn9bYO8fywgH989Lz9Rx9UFZq7OTyAvUTvsYfOzr7Vj0HNsMl5w9Li/M/i/37AIJYnZepJzYkmaHEN8uhblZ8J6d3Mnm976gJqeQ8jITI5OY97CK0idJtbgBuhz9bGzc2coGNdaavEFgqPGKdEpTDVNDfUa58TmDPuo8Zn+XQuXFvFaTxzitT6z2tpa8vLyxrqMYWGz2dBqR/av+af6ekmStFuW5bJTHS9Glk/D+vZaVOoA6vnXi6B8gbz+AJ809vFedSfra7rosnpQKiRmZBi4Y2Yai/LNpMSO7IQsr0umfmcnbfXBFSsGuo6F40gliVkx5M5KPDZyHH3aVSpaa5vYtO5DGgaakZDIj83i8mvmY5qSMqK1j3dev5fKnkq2tm3l4/aPqbXUAhCpjKTAWMDd+XeHRo2NauMYVysIgiAI50aE5VPw2x3YN20mJkO0YJwv15Cfj+p7eK+6kw9qu7C6fajDlFyRE8/VBWYW5I7sChayLNPbaqexsofGvT1Y2mXqqSE8Ukni5Bjy5hwLx6mnD8fHHd51kM0bP6LJ2UGYrGS6OZ+51y8gNnXiBr8WWwtb27aytX0rOzp24PQ5UUkqiuOL+da0bzE7aTZTDFPErniCIAjCRU+E5VOwV1Qge33o8nSQetlYl3PRGHR6+eBgF+sOdLKpoQe3NzhB7+qCBBYXJDBvspHIMOWIXV8OyHQ2DnK4socjlT1Ye91IEiRmx2Aukbj8mlKMqVoUZ7HFdSAQ4OCWfWzespWOoV4iCGN26jRm3zifaKN+xJ7DeOX0OtnZuZMtbVv4uP1jjtqOApAcncx1k65jTvIcZibMJDpcbLAiCIIgXFrGZVge66XjrGvfQKX2o154GyjEagZn0m11815NF+8d6OSTxj58AZkEXSS3laWypCCBGZmGEZ2g5/cFaKvvp3FvD0f29eK0DqFQSaTmGihdkkFGsZEoXTgVFRWY0nVncT4/+9bv4ONdn9DrH0RDJFdmX8bMGy4nUjdx1u6VZZn6/nq2tm9la9tW9nTvwRfwoVapKTOXcUfeHcxJmkO6Ll1MZhQEQRAuaeMyLI/l0nF+uwPH1o+JyXSJjUhOo6nXwXvVwQ1C9hwdAGCSUcPX501iSWECxcn6sxq9PV/eIT8t1RYOV3bTVNXHkMuHKkJJekEcWdPiSS+MI1x9bt/aXvcQu97eyicHdjEoO4iRNCwuuoLy6+aiGsX1nMdSv7ufbe3b2Noe7D0+vmLF5NjJ3JV3F3OS5zDdNJ1wpejhFwRBECaOcRmWx5L9ww+RvX50xYlgPvVSJRNR56CbV3a3sLaqg4OdNgAKk3U8vCiHJYUJZJuiR3SE0e3w0nygj8a9PRyt7sPnDRChUTGpxMikaSZSc2NRhZ97i4fb5uSTNz9iZ0MlDjwYlXpuLFtM8aIZKFUj1zIyHvgCPqp6qkKjxzV9NcjI6CP0zEqcxeyk2cxOmo1ZYx7rUgVBEARhzIiw/BmB7iYiYodQX718rEsZc/6ATEVdN6t3tLDxYBcBGcozYvne0nyuzjeTahj+tgRZlrFZ3PS12uk94c3a4wJAExNB3pwkJpUYSZoc84WT807H3jPIljc/ZG9LNR68JIcbWTrnGqbMK0ZxCbfetNvbQ+F4e8d27F47CklBsbGY+0ruY27SXPLj8kNbRwuCIAjC2ZAkiTvvvJN//vOfAPh8PhITE5k5cyZr167lxRdfZNeuXTz77LM89dRTPPPMMzQ1NWEymQCIjo7Gbg8uJ6tUKikqKsLr9aJSqbj77rt58MEHx+z3swjLnxGb7SLm6l6k4lvHupQx0zbg4n93tvC/u1roGHRjjI7gniuyWF6eSnqcZtiu4/cGsHQ46G21BUNxi52+NjseZ3BNXiTQx6uJT40mb1YCKXkGzOk6pAto8bC09rDlzY1UddXjk/xkRiUxb8HlTCrLHaZnNb64fC52de7i4/aP2dK2hSZrEwAJmgQWZywOTsxLnIku/Iv7uQVBEAThdDQaDQcOHMDlcqFWq1m/fj3JycmnPd5oNPKLX/yCp59++nOfU6vVVFZWAtDd3c0dd9yB1WrlBz/4wYjVfyYiLH9W7rVIEVqISR3rSkaV1x9g48FuXtpxlIr6HgDmTY7n+9fnc1WembALnKTnHfLTeXiQ3hY7vW02+lrt9Hc4CQSCm+KowhXEJUeTXWrCmBKNMVWLIUlDeOTwfIv6epz859f/oLa/ERnIiUnn8iVXkpyXPiznH096Xb180PwBHxz9gN1duxkKDBGhjKDMXMatObcyN3kumfpMMTFPEAThEvT0jqc5aDk4rOfMNeTy3Rnf/cLjrr32Wt5++22WLVvG6tWrWbFiBZs3bz7lsatWreLFF1/ku9/97hm3sjaZTDz33HOUl5fz1FNPjcnvLhGWP8tcMKF6lVssTl4+NorcbfNg0kZw/5XZ3FaWesFtFm6Hl+b9vTRW9nK0pg/fUAAAjT4cY6qW9CJjMBinRKM3RY3IpMC+pi7Wv/ouddYmFCgoNE1m3tIFxKcnDPu1xlKHvYMNRzewoXkDe7v3IiOTocvg9tzbmZM0h1JzKZGqyLEuUxAEQbiELV++nB/+8IcsXbqUqqoqVq1addqwHB0dzapVq/jNb37zhSPGkyZNwu/3093djdk8+vNoRFiegIZ8ATbUdrF6x1G2HOpFAuZPMbFiRhpXTom/oKXeHAMejuzrobGyh7a6AQIBmSh9OLmXJZJRbMSUrkWtHfnVFFz9DipeXseujmoAsrWpLP3yLejNsSN+7dHSbG1mffN6NjRvoLov+DxzYnO4r+Q+FqUtIismS4weC4IgTDBnMwI8UoqLi2lqamL16tVce+21X3j8Aw88QElJCY888sgoVHf+RFieQFosTv61/Siv7G6h1z5Eoj6Sb181mdvKUkmKUZ/3eQe6nMGd8ip76DpiBUBvUjN1YSqTSuIxZ1xYn/G58Hu8bF+ziS0Hd+CUPOTEpLN42XXsP1xz0QdlWZY5NHCIDc0bWH90PQ39DQAUxhXynenfYWH6QtJ1l15biSAIgnDxuOGGG3jkkUeoqKigr6/vjMfGxMRwxx138Pvf//6MxzU2NqJUKkOTAUebCMuXuEBA5qP6Hv6+rYmK+h4UksSCXBN3zEjj8px4lOcRYmVZprfFHgrIlnYHAPFpWmbekElmSTyGRM2ojmrKAZna9Xv44JMK+mQb5kgDt163jMziycEDDteMWi3DSZZlaiw1bGgOtlg0WZuQkJhmmsaj5Y+yMG0hidGJY12mIAiCIADBXuSYmBiKioqoqKj4wuMfeughysvL8fl8p/x8T08P9957L/fff/+Y/bV0XIblsd7B71LQ7xjif3e18M/tzbRYXBijI/jWldmsmJlGov7cR5F9Q346GwdpquqjsbIHm+XTraTn3jqZzBIjurjzH52+EO27Gnn/vfdp8nYSrVBz49xrKFkw46JtQQjIAap6qljfvJ4Pjn5Am70NpaSkLKGMu/LuYkHaAuKj4se6TEEQBEH4nJSUFB544IGzPt5oNHLzzTfzq1/9KnSfy+WipKQktHTcl7/8ZR566KGRKPesjMuwPJY7+F3MhnwBKuq6eb2yjQ213Qz5AszINPDdJblcnZ9AuOrse5H93gBdTYO01g3QVtdP55FBAj45uJV0noGy6zLILDaOSv/x6Qw09rBxzTr22xpRSUouz7+MuTcuIDzi4tthzhfwsadrD+ub17Px6Ea6Xd2oFCpmJc7inuJ7mJ86n9jIi7uNRBAEQbh0HV8j+UTz589n/vz5AKxcuZKVK1cC8NRTT5103C9/+Ut++ctfhm77/f6RKvO8jMuwLJy9QEBmV3M/r+1t4539HQy6vMRpwrljRhorZqQxJUF7Vufx+wN0N9loq+unrb6fjsOD+L0BkCA+VUvx/BSSp8SSNDlm2JZzO18ei5MtL69nR+d+hiQfRSm5LLr1WrQxF9dawV6/l+2d29nQvIGNRzfS7+knUhnJ3OS5LExfyOUpl6MNP7vXTxAEQRCEkSHC8kWqvsvG63vbeKOynbYBF+owJYsLzNw4LZm52cYvXBc54A/Qc9ROW30/bXX9tB8exOcJ/k8uLllDwdykUDiO1ISNxlP6Qn6Xl8rXPuajuk+wSi7S9Ilc86WlJKafftHz8cbtc/Nx+8dsaN5ARWsFtiEbmjANl6dczqL0RcxJmkNU2PDvjCgIgiAIwvkRYfki0jno5s19bby2t53aDitKhcTcbCP/tXgKi/LNaCLO/HI6Bj0c2tVNy0EL7Q0DeN3BcBybEEXuZQkk58SSnBMzpq0VpyIHZBrX72fDtg/poJ/YCB3Lr1nKlJL8i6Iv2el1sqltExuaN7CpdRMunwtduI4FqQtYlL6Iy5IuI0IZMdZlCoIgCIJwCiIsj3OyLLO+posXP25iW2MfsgxTU2N46vp8ritOIl575pDl9wdo3t9H7dZ2mqstyAEZvUnN5HIzKTmxJOXEoNGP36A2ZHGx4YU32WmtJUIZzuJZC5mxYBZKpXKsSzsjf8DPx+0fs6ZhDZvbNuPxezBEGlg6aSkL0xdSnlBOmGJ8jNgLgiAIgnB6IiyPYzXtVn60toZtjX2kGaJ4YMFkbpqWTKZRc8bHBQIy/R0ODm7roG57Jy6blyh9ONMWpZE7K4HYhDM/frzo/KSR19e9RSf95CVP5oYv34JaPTYrbpytLkcXrx16jTUNa+hwdGCINLAsZxkL0xYyzTQNpWJ8h3xBEARBEE4mwvI41Gv38Iv363hpZwt6dRg/vLGAO2akoVIq8HsDWHtd2Ps9OAY8n74fcIc+dgwOIQdkFAqJjGIjebMTSSswoLiAnflGU8DjY/c/PmJDyzb8ksz1C6+jdG75WJd1Wv6An63tW/lP/X/Y1LqJgBxgVuIsHil7hCtTryRMKUaQBUEQBOFiJcLyGJEDMkMePx6HF4/Th8fpxW4b4sP9nWyt6UHpk7nfZGCKQUNg5wCvbuzGMeDBZfN+7lyqCCXRMRFoYiJInhKLJiYCXVwkmVPjidKNr/7jL2I/3Mc7q9+kxteMKTqO276yHKNpfK4p3OnoZE3DGtY0rKHL2UVcZByrCldxy+RbSNWmjnV5giAIgjBqlEolRUVFobWR7777bh588EEUik8H6r7zne/wn//8h5aWltD9L774IqtWraKyspLi4mIACgsLWbt2LRkZGdjtdh5++GE2bNhATEwMWq2Wp59+mpkzZ47acxNheYT4fQFsFje2XjfWPhfWXje2PhfWPjfWPjdu2xCyfOrHzkKJpIAIi48Bt4OIqDA0+ghMGbpQKI6OiUATG3wfrlZdFBPdzkT2yzS9XcWbu9bTr7Azo2A6V998LSrV+PoW9QV8bG7dzCsNr7ClbQuyLDM7eTaPzXiMK1KvEH3IgiAIwoSkVquprKwEoLu7mzvuuAOr1coPfvADAAKBAK+99hqpqal89NFHXHnllaHHpqSk8JOf/ISXX375c+f9+te/TmZmJg0NDSgUCo4cOUJNzejuyju+kshFIhCQcdu9OK0enINDOAY9WPvc2PrcWHtd2Prc2Ac8cEIYVigkouMigyO+RXFE6SOIiFLRO+TltQMd7Ou2kRCn5t5Fk7m80ExYhPKiD8Bny9vrZMv/e58t1irCw8JYcctyphTkjnVZJ2m3t7OmYQ2vNbxGt6ubeHU8Xy/6OrdMvoXk6Itn6TpBEATh0tb5P/+Dp/bgsJ4zIi+XhCeeOOvjTSYTzz33HOXl5Tz11FNIkkRFRQUFBQXcfvvtrF69+qSwvHTpUjZt2kRdXR1JSUmh+w8fPsz27dv517/+FRqJzszMJDMzEwhuZvK3v/0NCIbq73znO8PxdD9HhOXPcAx6GOx24bQGQ7DTOhR8GxwKhWPXqUaFJYiOiUAbF0nylFh0cZHojGq0x95rYiJQKD4Nvz22YF/yy7taiFGH8dCyfFaUp6K6SPqKh4Msy1g+aWHturc5InWRYU7hS1++Ha12fGzE4Q142dSyiVcaXmFr21YA5ibP5cmcJ7k85XJUCvHPRxAEQRBOZdKkSfj9frq7uzGbzaxevZoVK1Zw44038sQTT+D1egkLC/41VqFQ8Oijj/I///M/PPvss6FzVFdXU1JScsoVsHbv3s0LL7zA9u3bkWWZmTNncsUVVzBt2rRhfy7it/1nHNjUxq63m0K3FQoJtS6cKF04mpgI4tO0wY/1EUQduz9KH050bCTKs9hO2jXk54WPj/CHDw/j9vr52pxMvnXVZPTqifXn+4DTS/W/t/Fey8c4JQ9Xzb2SOQvmndTbNFZaba3BUeRDr9Hr6sUUZeKeqfdwc/bNJEUnffEJBEEQBGGMnMsI8GgZGhrinXfe4Ze//CVarZaZM2fy3nvvsXTp0tAxd9xxBz/5yU9oamo6q3Nu2bKFm2++GY0muMLXLbfcwubNm0VYHg055WaSsmKI0geDcKQmDElx4e0QPn+A/+xu5dcb6umyeliYZ+KJa/OYFB89DFVfXJwNFja+9C67fQ3o1FpW3XkXKakpY1qTN+CloqWCV+pfYVv7NiRJYl7yPJblLGNu8lwxiiwIgiAI56CxsRGlUonJZGLt2rUMDAxQVFQEgNPpRK1WnxSWVSoVDz/8ML/+9a9D9xUUFLBv3z78fv+Y7q8wLhOAJEnXA9dnZ2eP+rVjEzTDug6xLMu8V93JM+/V0djjYHpaDL9bMZ0ZmYZhu8bFQvYFaF1bw9t7NtCpGCA/O5cbbr2JyMjIMaupxdrCqw2v8vqh1+lz95GgSeC+qfdx8+SbSdAkjFldgiAIgnCx6unp4d577+X+++9HkiRWr17N888/z4oVKwBwOBxkZmbidDpPetzKlSv56U9/isPhACArK4uysjK+//3v86Mf/QhJkmhqaqK6upp58+axcuVKHnvsMWRZ5rXXXuMf//jHiDyfcRmWZVl+C3irrKzs/4x1LRdi2+E+nl53kMqWAbJN0fz5y6VcnW+eMBP3TuTtcbL77xV8aN1DQAU3XnsDJaXTxuRr4fV7+aDlA16tf5VPOj5BKSm5POVyluUsY07SHLFxiCAIgiCcI5fLRUlJSWjpuC9/+cs89NBDOJ1O1q1bx5/+9KfQsRqNhrlz5/LWW2+ddI7w8HDuvfdevvvd74bue/7553n44YfJzs5GrVZjNBr52c9+xvTp01m5ciUzZswAghP8RqIFA8ZpWL7Y1bRbeea9g1TU9ZCoj+SZLxVzy/TkCTV570TWPZ2898Y7VEtHMcfGc+tdt2M0Gke9jmZrM6/Wv8obh9/A4raQpEni/pL7uSn7Jswa86jXIwiCIAiXCr/ff8r7o6KisFgsn7t/zZo1oY9XrlwZ+vi+++7j0UcfDd3W6XT85S9/OeW5H3roIR566KHzrPjsibA8jFosTn7xfh1v7GtHFxnG49fk8pXZGUSGTcyRStkXoHnNPtYe2EivwsaMkjKuXrpkVNdOlmWZTa2b+G3nb2l4rQGlpGR+6nyW5SxjVuIsMYosCIIgCMIZibA8DPrsHn638RD/2t6MUiFx7xVZ3HtF1oRb4eJEPoubHS98wIfWPSjClNx+y23kFeSPag11ljp+tvNnbO/cjkFp4IFpD3BT9k3ER43PHQEFQRAEQRh/RFi+QFsaern3n7txef3cVpbCt6/KIUE/dhPWxgPb/m7efXUtNRwl0WDmtruXExsbO2rX73X18uzeZ1nTsAZdhI7HZzyOudPMVcVXjVoNgiAIgiBcGkRYvgAf1ffwjb/vItOo4dk7ppNtmnjLwJ1I9ss0v1nFW3s30KewMbOknEVLF49a24Xb5+YfNf/g+f3PMxQY4q78u7in+B70EXoquipGpQZBEARBEC4tIiyfpw8PdnPPP3aTbYrmn1+fiUETPtYljSn/oIftL3zAh/27UYQpWX7L7eQW5I3KtWVZZl3TOn61+1d0ODpYkLqAh8oeIl2XPirXFwRBEATh0iXC8nnYUNPFff/azZQELf/82kxioiZ2ULYf7OGdl9+iRg62Xdz+lRXExMSMyrX39ezjmZ3PUNVTRa4hlx/P+TEzEmeMyrUFQRAEQbj0ibB8jtYd6OT+f++hIEnH31fNRB81cSfxyQGZo28f4M2d7wfbLqaWsej60VntosPewa/2/Ip3j7yLUW3kh7N/yA1ZN4jVLQRBEARhDCiVytAOfQDLly/nscceY+3atXzve98jEAjg9Xr59re/zT333ENdXR333HMPAwMDeDwe5s2bx3PPPTeGz+D0RFg+B29XdfDAS3spTtHz/1bNQBc5cYOy3z7E9hc28GHvbpSj2Hbh9Dp5fv/z/L3m7wB8o/gbfK3wa0SFRY34tQVBEARBODW1Wk1lZeVJ93m9Xr7xjW+wY8cOUlJS8Hg8NDU1AfDAAw/w4IMPcuONNwKwf//+0S75rImwfJbe3NfOgy9XMi01hhe+Wo52Agdlx6E+3v73G9QEjpIUa+a2lSPfduEP+Hnj8Bv8bu/v6HX1ct2k6/j2tG+TGJ04otcVBEEQhIvJ5v+tp7fFPqznNKZGM++2nHN+nM1mw+fzERcXB0BERARTpkwBoKOjg5SUlNCxJ45KjzciLJ+F1/a28vD/7qMsw8ALK8vRREzML5ssy7S8V8MbH79Ln8LOZcXlLLpxCUrlyLY+7OzcyTM7n+Gg5SAl8SX85srfUBxfPKLXFARBEATh7B3f7vq4xx9/nNtvv50bbriB9PR0rrrqKpYuXcqKFStQKBQ8+OCDLFiwgNmzZ3P11Vfz1a9+ddTmO52riZn6zsEru1v5r1f2cVlmHH9dWUZU+MT8ksneANv/uoEPOrajUB1ruygc2baLdns7v9j1C95vfp9ETSI/u/xnLM5YjCRJI3pdQRAEQbhYnc8I8HA4VRsGwPPPP8/+/fvZsGEDP//5z1m/fj0vvvgiX/3qV1m8eDHr1q3jjTfe4M9//jP79u0bg8q/2MRMfmfp5Z1HeWzNfuZmG3nuy2Wowyfm5LGA28eHf3iLzdZ9/P/t3Xl8VfWd//HX9yY3yc2+QggggbDILhDBXRYXRHFv1VoVbatO63Sz09p2puN0amsdx3Fs7c8yrcO044hW3LC4AIWiVtn3PWHNvu/bzb3f3x/3EgIkEOAm9yZ5Px+P+7jnfO85534uJxfefPM93zMoYQB3PfSlbv3fX4O7gZd3vMyinYswGL5x0TdYMH4BUeH9+2YvIiIivdHEiROZOHEi9913H8OHD2fRokUAZGRk8NBDD/HQQw8xYcIEduzYwejRwQn7p6Ow3IlX1h7mx2/t4OrRafz2vmlEOftnUG6taeaD37zJhqa9jEofzhe/8iWczu4Zr22tZdnBZTy38TlKGkqYN3we35n2HdJj0rvl/URERKT71NXVsWHDBmbOnAnAli1bGDbMdw+EDz74gDlz5uB0OikqKqK8vJzBgwcHsdrOKSx34A+fHeIn7+xk9oUD+M29U/ttUHaXNfDOS6+zo/UQE4eP5dYv39lt45N3lu3k6XVPs6V0C+NSxvHs1c8yZcCUbnkvERERCayTxyzPnTuXH//4xzzzzDM88sgjuFwuYmJi2nqVP/roI771rW8RFeX7rfG//du/kZ6eTm1tbTDKPy2F5ZMcC8rXjhvIr780hcjw/hmUG/OqWfL7xeTYQqaPn8bcO27E4XAE/H3KGsv4z03/yds5b5MSlcJPL/spt4y8BYcJ/HuJiIhI9/B4PB22L1u2rMP25557jueee647SwoYheWTjBkYx+1TBvP0HZOICO+fga1uXymvvfIaR00ZM2dcxdVzZwX8oroWTwv/u/t/+e3W39LibeHBCQ/y8MSHiY2IDej7iIiIiJwPheWTzBiRwowRKcEuI2gqNxew+O3XKTZV3DB7LjOuuiSgx7fWsuroKp7d8CxHa48yc+hMvpf9PYbFDwvo+4iIiIgEQo+FZWPMCODHQIK19s6eel/pupJPDrD4ozepctRz+/zbmDRtckCPn1OZwy/X/5LPCz9nRMIIfnvNb7ls8GUBfQ8RERGRQOpSWDbGvAzcBJRYaye0a58L/CcQBvzOWvt0Z8ew1h4AvmKMeeP8SpZAs9aS98Fu/vTZUhrCWrj7rq8cPzcAACAASURBVHsYfWHgpm7xeD28tO0l/mvbfxHtjOaJ6U/wxTFfxOnov3dBFBERkd6hqz3Li4BfA3841mCMCQNeBK4F8oD1xph38QXnX5y0/0PW2pLzrlYCznotB97YxJKdH+EJt9x///1ckBm4IRElDSX8YM0P2FC8gfkj5vMPF/8DSVFJATu+iIiISHfqUli21q4xxmSe1DwdyPH3GGOMWQzcYq39Bb5eaAlx1uNl9x8+451DqwiPcPLQVxYwMH1gwI7/af6n/OiTH9HY2sjPLv8Zt4y8JWDHFhEREekJxlrbtQ19Yfm9Y8MwjDF3AnOttV/1r98HzLDWPtbJ/inAU/h6on/nD9Udbfcw8DDAwIEDpy1evPhsPs9ZqaurIza2f86+YFrBbqjik8btRDojmDB1Eq5oV0CO7bEe/lz1Z5bXLGeQcxAPpj7IoIhBATn2uerP57q/0bnuP3Su+w+d69NLSEhg5MiRQa0hMTGR8ePH09raypgxY3jppZeIjo6muLiYH/zgB2zatInExETS0tJ4+umnGTVqFAAvvvgiTz75JDk5OSQkJODxeLrtng7H5OTkUF1dfULbrFmzNlprszvavscu8LPWlgOPdmG7hcBCgOzsbHvsri/dYfXq1XTn8UOVp97Nut9+xPLGraTEJ3P/wwuIi4sLyLGL6ov4/prvs7lmM3eMuoMfTP8BrvDAhPDz0V/PdX+kc91/6Fz3HzrXp7d79+6A/Tt+rlwuF9u2bQPg3nvv5ZVXXuE73/kO1113HQ888ABLliwBYOvWrdTU1LTV+9Zbb3HxxRezfPlyHnzwQWpra7v9s0RFRTFlStdvfHY+YTkfGNpufYi/TUJYa1Uza156j48bt5OROogvf/V+XK7AhNk1eWv40Sc/wu1x8/SVT3PjiBsDclwRERHpmlWLFlJy+EBAjzlg2AhmLXi4y9tfeeWVbNu2jVWrVuF0Onn00eN9pZMnH59pKzc3l7q6On7zm9/w1FNP8eCDDwa07kA5n7turAdGGWOGG2MigLuBdwNRlDFmvjFm4cld5HJ+WqubWf6rN1nTtJ3hQzJ54JEHAxKU3R43z65/lm+s/Abp0em8dtNrCsoiIiL9UGtrK++//z4TJ05kx44dTJs2rdNtFy9ezN13382VV17J3r17KS4u7sFKu66rU8e9CswEUo0xecA/W2t/b4x5DPgQ3wwYL1trdwaiKGvtUmBpdnb21wJxPAHr9rJm4TLWevYyLutCbr/nTsLDz38UTn5dPt//6/fZVraNu8bcxT9c/A9EhkUGoGIRERE5W2fTAxxIjY2NXHTRRYCvZ/krX/kKL7300mn3efXVV3nrrbdwOBzccccd/OlPf+KBBx7oiXLPSldnw7ink/ZlQMc3/ZaQYa1lx/9+ysd12xgxaBh33vtFHI7zv5X3yiMr+adP/wlrLc9e/SzXZ14fgGpFRESkt3G5XGzZsuWEtvHjx/PGGx3fXmP79u3s37+fa6+9FoCWlhaGDx8ekmH5/BOThLz8lXv586E1JEbH84UH7j7voNziaeHpdU/z7VXfZmjcUF6/6XUFZRERETnB7NmzaW5uZuHChW1t27Zt4+OPP+bVV1/lySef5NChQxw6dIiCggIKCgo4cuRIECvuWEiGZY1ZDpyaPSW8ueY9vGFw70P3nfcY5aM1R7nv/ft4ZfcrfHnsl/njDX9kaPzQM+8oIiIi/YoxhrfeeosVK1aQlZXF+PHj+eEPf0h6ejqLFy/mtttuO2H72267rW3WjFDSY1PHnQ2NWQ4Md3kjby1eQqWjjnvuvIfUtNTzOt6Hhz7kyb89iTGG52c9z5wL5gSoUhEREenN6urqOmzPyMjg9ddfP6X9wIFTZ+x47rnnqK2tDXht5yskw7KcP2+Lh4/+620OUsw1l89i9Lgx53ysFk8Lz254llf3vMqk1Ek8c/UzDI4dHMBqRUREREKTwnIfZK1l/aK/sL5pLxMyL+Tya64652Pl1ebxvb9+j53lO7l/3P18e+q3cYY5A1itiIiISOhSWO6DDn6wneX5n5Men8Yt996BMeacjrPqyCp+/OmPwcLzM59nzjANuxAREZH+JSTDsjFmPjA/2Pc5742qdhfz9ufvExHu5Etfuw+n8+x7gd1eNy9seoFFOxcxNnks/z7z3xkap4v4REREpP8JydkwrLVLrbUPJyQkBLuUXsVd2cTbry2h1jTxhXu+SHx8/Fkfo6i+iIc+eIhFOxdx15i7+OO8Pyooi4iISL8Vkj3Lcvas28vK/3qXQ5RwzeWzGD5yxFkf45P8T/jhxz+kxdPCM1c9ww3Db+iGSkVERER6j5DsWZazt/3VT1lbv4sLh4w86wv6Wr2tvLDpBb6+4uukRaex+KbFCsoiIiJyVoqKirj77rvJyspi2rRpzJs3j3379uFyuZgyZQpjx45l+vTpLFq0KNilnhX1LPcBhWtyWJa7hqToBG677wtndUFfWWMZ31/zfdYXref2UbfzxPQncIWf341LREREpH+x1nLbbbfxwAMPsHjxYgC2bt1KcXExWVlZbN68GfDNr3z77bdjreXBBx8MZsldFpJhWRf4dV3DwUreWrGU1jAvdy+4l8jIyC7vu65wHd9f833q3fX87PKfccvIW7qxUhEREeluVUtzaSmoD+gxIzJiSJyfddptVq1ahdPp5NFHH21rmzx5MocOHTphuxEjRvDcc8/x+OOP95qwHJLDMHSBX9d4altY9se3KXFUc/NNNzNg4IAu7ee1XhZuW8jXln+N+Mh4/u/G/1NQFhERkXO2Y8cOpk2b1qVtp06dyp49e7q5osAJyZ5lOTPrsXz+++Xs8B7m4gnTmDRtcpf2q2yq5Icf/5BPCz5l3vB5/POl/0y0M7qbqxUREZGecKYe4FBgrQ12CWdFYbmXyn1zE6sqNzI4JZ25t83r0j6Haw7zlQ+/QmVTJT+59CfcOerOc75hiYiIiMgx48eP54033ujStps3b2bs2LHdXFHghOQwDDm9yg15LN2+kghnBHc/eC9hYWFn3Ke0oZRHlj9Ci6eFP877I18YfXYXAoqIiIh0Zvbs2TQ3N7Nw4cK2tm3btnH06NETtjt06BDf+973+Pu///ueLvGcqWe5l2kprOOdpe9S42jkgXvuJy4u7oz71LbU8uiKR6loquDl619mXMq4HqhURERE+gtjDG+99Rbf/va3+eUvf0lUVBSZmZk8//zz5ObmMmXKFJqamoiLi+Ob3/wmCxYsCHbJXRaSYVmzYXTM29jKX/77PQ6ZEq65ag6ZWcPPuE+zp5lv/uWbHKg+wIuzX2RC6oQeqFRERET6m4yMDF5//fVT2hsbG4NQTeCE5DAMzYZxKuu1bP+fT1jbvJcxw0Zy+awrzriPx+vhB2t+wIbiDTx1+VNcNviyHqhUREREpO8IybAspyp8fw8fFP6NpNh4brvnzBfmWWv52dqfsfLISp6Y/gTzRnTtIkAREREROU5huReo213KO2s/wO3wcNf9XyIqKuqM+7y45UXe2PcGX534Ve4de28PVCkiIiLS9ygshzjb6mXlkg8pdlRzyy03M3DgwDPuszR3Kb/d9ltuG3kb35zyzR6oUkRERKRvUlgOcYVrctjmPsC4YWOYeNGZbzySW5XLv37+r2QPzOYnl/5E08OJiIiInAeF5RDmbfaw+uO/Yg1ce+vcM27f4G7g8dWP4wp38cxVzxDuCMnJTkRERER6DYXlEHbwg23s8+aTPWEKSUlJp93WWstTa5/iQPUBfnnVL0mLTuuhKkVERKS/i42NPWF90aJFPPbYY6fd59133+Xpp5/uzrICIiS7HjXPMnjqWli96VMiw5zMnDfnjNu/nfM27+a+y99N/jsuGXRJD1QoIiIicu5uvvlmbr755mCXcUYhGZattUuBpdnZ2V8Ldi3Bsuud9Rw1ZcyecTXR0dGn3XZf5T6eWvsUMwbN4JFJj/RQhSIiIhJq3n//fYqKigJ6zPT0dG644YZz3r+0tJRHH32UI0eOAPD8889z+eWXs2jRIjZs2MCvf/1rFixYQHx8PGvXrqW0tJRnnnmGO++8M1Af4byEZFju79zljazZu5bYyGgunX36m4/Uu+t5fPXjxEXE8fSVTxPmCOuhKkVERER8Ghsbueiii9rWKyoq2nqNv/Wtb/Gd73yHK664giNHjnD99deze/fuU45RWFjIRx99RH5+PjfffLPCsnRu45ufUOqo4eZZN+J0OjvdzlrLTz/7KUdqj/C7635Hqiu1B6sUERGRUHM+PcDnw+VysWXLlrb1Y73GACtWrGDXrl1tr9XU1FBXV3fKMW699VYcDgfjxo2juLi4+4vuIoXlENOYV8Pf8jaRGp3ERZdMO+22b+x/g2UHl/HYRY9xcfrFPVShiIiISNd5vV4+//zzM95ULTIysm3ZWtvdZXWZZsMIMZ8t+Qs1ppFr512Hw9H56dlTsYen1z7NZRmX8bVJ/XZot4iIiIS46667jl/96ldt6+17oHsDheUQUrOnhHUVOxmSOIjREy7sdLu6ljoeX/04iZGJ/OLKX+AwOo0iIiISml544QU2bNjApEmTGDduHC+99FKwSzorGoYRIqy1rHl3JU3GzfW3zuv0znvWWp787Eny6/L5/fW/JzkquYcrFRERETnRyWOQFyxYwIIFCwBITU3ltddeO2Wf9tssWrQIgNra2g6PF0zqkgwRZRuPsKV+P2MGjWBo5tBOt3tt72t8eOhDHpvyGNMGnn5Ms4iIiIicH4XlEGA9llUfrsRrLNfdfmOn2+0s38kz65/hysFX8tCEh3qwQhEREZH+KSTDsjFmvjFmYXV1dbBL6RF5a/ayu+UoF2VNICUtpcNtaltq+d7q75EclcxTVzylccoiIiLSJpRmjwhl5/LnFJKJy1q71Fr7cEJCQrBL6XbW7WHlx6sId4Qx+7brO97GWn7y6U8oqi/i2aufJSkqqYerFBERkVAVFRVFeXm5AvMZWGspLy8/4xR2J9MFfkG294MtHPIWc+XkS4mNje1wm9f2vsaKIyt4fNrjXDTgog63ERERkf5pyJAh5OXlUVpaGuxSzltTU9NZh9mzERUVxZAhQ85qH4XlIPI0uFm18WOiHVFceeOsDrcpqi/iPzb+B5dnXM4D4x/o4QpFREQk1DmdToYPHx7sMgJi9erVTJkyJdhlnCAkh2H0F1vf/hvFVHHVZVcSERHR4TbPrH8Gj/Xwj5f8Y6fTyYmIiIhI91BYDhJ3VSNr9q4jMSKOi2dd0uE2a/LWsPzwch6d/ChD4s7uVwYiIiIicv4UloPk8z+tpsrUc8011xAWFnbK642tjfx87c8ZkTCCB8Zp+IWIiIhIMGjMchDUF1TxWd5mBsWmMf7iSR1us3DbQvLr8vnv6/8bZ5izhysUEREREVDPclB8vGQlDaaF6+fP7XAcck5lDot2LOKWrFvITs8OQoUiIiIiAgrLPa6xrJZNZbsZkTSUzAuzTnnda7386+f/SkxEDN/N/m4QKhQRERGRYxSWe9i2j9bTYlq5dOblHb7+Ts47bCrZxHenfZfkqOQerk5ERERE2lNY7kHWa9mas5P4sGiyJo4+5fXKpkqe2/gcUwdM5daRtwahQhERERFpT2G5BxVvOUyBt5xJoyficJz6R/8fG/+DupY6/vGSf8RhdGpEREREgi0kE5kxZr4xZmF1dXWwSwmojZ+sA+Diay899bXijbyV8xb3j7+fUUmjero0EREREelASIZla+1Sa+3DCQkJwS4lYNy1zewqz2VYfAYJyYknvuZx87PPf0ZGTAaPTHokSBWKiIiIyMlCMiz3RbtXbqbeNDNtxqlTwf1h1x/IqcrhRzN+RLQzOgjViYiIiEhHFJZ7gLWWLTu2EmUiGDfjxJuQ5Nfl89LWl5hzwRyuHnp1kCoUERERkY4oLPeAqn3FHHIXMX7YGMLDj9800VrLz9f+HGMMT0x/IogVioiIiEhHFJZ7wKZV6/Aay8XXnHhh31+O/IU1eWv4xkXfID0mPUjViYiIiEhnFJa7mafJzY7CvQx0pZA+JKOtvd5dz8/X/ZwxSWO4d+y9QaxQRERERDqjsNzNDn68i0pTz5QpU05of3HLi5Q2lPJPl/4T4Y7wTvYWERERkWBSWO5mmzZuJhwHk6+c1ta2u3w3r+x+hS+M/gKT0yYHsToREREROR2F5W5Un1fJ/sajjE7PwuVytbX/fO3PSYxM5JtTvxnE6kRERETkTBSWu9HWFRtwGw/ZV89oa9tZvpMtpVt4eNLDJET2nZuuiIiIiPRFCsvdxLZ62X5oFwnOWIZfmNXWvmTfEqLCopifNT+I1YmIiIhIVygsd5P8dbkUUslF4yZhjAGgwd3AsoPLuC7zOuIj4oNcoYiIiIicicJyN9n02QYMMG32JW1tHx76kHp3PXeMuiN4hYmIiIhIlyksd4OWsgb21BxkeNJQ4hOO9yC/sf8NRiSMYMqAKafZW0RERERChcJyN9i5chMNpoVpl1/c1ra/cj/bSrdx+6jb24ZliIiIiEhoU1gOMOu1bN27HZcjkgunjG9rf3P/m4Q7wrk56+YgViciIiIiZ0NhOcAqtuVz2FPChKyxhIWFAdDsaWbpgaXMuWAOSVFJQa5QRERERLpKYTnANn28Dmss0+dc1ta24vAKqpurdWGfiIiISC+jsBxArbXN7CjLYVBMGmnpA9ral+xfwuDYwcwYNOM0e4uIiIhIqFFYDqCc1TuoNg1MvXhqW9vhmsOsL1rPHaPuwGH0xy0iIiLSm4T31BsZY24FbgTigd9baz/qqffuCdZatmzbQrgJY9Klx8Pykv1LCDNh3Dry1iBWJyIiIiLnoktdncaYl40xJcaYHSe1zzXG7DXG5BhjnjjdMay1b1trvwY8Ctx17iWHprrccnJaChg7eBSRkZEAuL1u3sl5h6uGXEVadFqQKxQRERGRs9XVnuVFwK+BPxxrMMaEAS8C1wJ5wHpjzLtAGPCLk/Z/yFpb4l/+R/9+fcqWv6yn1XjInnX8jn1/PfpXKpoqdGGfiIiISC9lrLVd29CYTOA9a+0E//qlwJPW2uv96z8EsNaeHJSP7W+Ap4Hl1toVp3mfh4GHAQYOHDht8eLFXf0sZ62uro7Y2NjzPo5phZ1rNtPodDP1iovbbjrym+LfUOgu5MnBTxJmws77feTcBepcS+jTue4/dK77D53r/iNY53rWrFkbrbXZHb12PmOWBwNH263nAaeb7uHvgWuABGPMSGvtSx1tZK1dCCwEyM7OtjNnzjyPEk9v9erVBOL4h1fuYpWjmllTr+TqWbMAKKgrYM+SPTw86WHmTJlz3u8h5ydQ51pCn851/6Fz3X/oXPcfoXiue+wCP2vtC8ALPfV+PWnzhk04MEy9anpb21s5bwFw+6jbg1WWiIiIiJyn85nLLB8Y2m59iL+tX2kqqmVPw2FGpF1AXFwcAB6vh7f2v8VlGZeREZsR5ApFRERE5FydT1heD4wyxgw3xkQAdwPvBqIoY8x8Y8zC6urqQByuWx3enkOTcTN56kVtbZ8WfEpxQzF3jNaFfSIiIiK9WVenjnsV+AwYY4zJM8Z8xVrbCjwGfAjsBl631u4MRFHW2qXW2ocTEhICcbhuVXCkAIBhY7Pa2pbsW0JyVDIzh8wMUlUiIiIiEghdGrNsrb2nk/ZlwLKAVtTLlJSXEGUiiEvwDcEobSjlr3l/5f5x9+MMcwa5OhERERE5H7r/8nmw1lLaUElaTHLbdHHv5L6Dx3p0YZ+IiIhIHxCSYbm3jFl2VzVRaesYmDoAAK/1smTfErIHZpOZkBnc4kRERETkvIVkWO4tY5aL9h7FY7ykDx0EwLqideTV5enCPhEREZE+IiTDcm9ReNA3U97g0cMA34V98RHxXDvs2mCWJSIiIiIBorB8HopLinHgYEDGQCqbKll5ZCXzs+YTGRYZ7NJEREREJABCMiz3ljHLJTVlpETGExYWxtLcpbi9bl3YJyIiInKWrNdLfVUlzbWhl/167HbXZ8NauxRYmp2d/bVg19IZb3MrZa3VZKVmYq1lyf4lTEqdxOik0cEuTURERCRkeL0eGqqqqK0oo668nNryUmoryqktL6Ouosz/XI7X4yF+aCbMvyXYJZ8gJMNyb1CZW0KTcTNwUDpbSrdwoPoA/3LZvwS7LBEREZEe4/V6qK+qPB6Cy8v9odgXgmsryqivrMDr8ZywX7gzgtiUFOKSUxly4XhiU1KJS07laFl5kD5J5xSWz1FBzlEAMkYO5b/3/Q/R4dHMzZwb5KpEREREAsPr8QXh2vKyE3qBj4XguvJy6irLsV7vCfuFOyOIS00lNjmVoeMmEpfiW45LSfUvp+CKi2+7R0V7VatX99Cn6zqF5XNUVFAIQPyQRD7a8hE3Zd1EtDM6yFWJiIhIqPF6PbibmnE3N+FuaqSlqcm/fPy5panxhHV3cxPWWrAW639gASzW6/Uvtn/NtrWd0I71N3l9r8GJ+xxr83rbtm1tbvb3CFf69msnPCLSH3pTuGDCJH8ITiEuJY3Y5BTiUlKJio3rMAj3ViEZlo0x84H5I0eODHYpnSquLCU2zMXOmp00eZq4cfiNwS5JREREAsDT6qahpprGmhpaGhvawusJIbep8XhbW3sj7qZmX/BtbvavN9Hqbjmr93dGuQiPiMDhcPhCp/9hMBiHAQzGcEqbbx2Mwzd/w7F9Tbv9fZs5Ttj/WJtvW3C6XAybOIW4lBRfGE71DZGITUklKia2TwXhrgjJsBzqF/hZj6WsqYoBiSlsLd1KuAlnQuqEYJclIiIiHXA3NfnDbzUNtb4Q3LZeU01jbQ2NNcfbWxobznxQY3BGRhERFYUzMgqn/zkiOprY5BSckZE4o1zH26P825zcFulrO7YcHhHRFnYlNIRkWA51jcU1VFPP2IEXsrrsb4xJHkNUeFSwyxIREenzrLU019e3hd3G2pp2y9U01NQcD8E1NTTW1tDa0tzhsRxh4UTHx+OKT8AVn0B82kCi4xNwxcf7nuPiiYiO8Qdd1wmhNzwist/1sPZXCsvnoHDvEayBgRcMYsfhHdw28rZglyQiIhLyrLW+oQuNjbQ0Nvif/ctNjSe2NzW0rRfl5XH4z2+0heOTZ1Y4xhkZhSs+gej4eGISEkkdOsy/3j4AH1+OcEUr8MoZKSyfg8IjBQC4B3ppzG1kUtqkIFckIiISWNZavJ5WPG43rW43rS0tvovTTgm4DSeuN3QUfBvbno9dZHY6xjiIcLlwulxEuqLxelpJGHQBg0aObhd+fT2/x4KwKz4BZ4TuoCuBp7B8DopLS3ASxgHPQQCFZRERCTivx9PWw+puamoLmx632xdgW91ty55jy62tx5f97a0dbXPSMbytrR0e72w4wsKIcEUT4XL5nqNcRMXGEZ820NcW5SIi2tfets2x9pP2C488cYjD6tWrmTlzZoD/hEW6RmH5HJTWVZDqSmJb2TaSo5IZEjsk2CWJiEiQeVpb/VOAtRte0NR4vDfW/9x+G/ex9qYm3O2GIpzLDArthTmdhIU7fc9OJ+HHlsOdhDnDCQt3EhEdjSs8/IRtw0/aL6zdfuERzpMCb/QJgTfM6dSQBumTQjIsh/LUca01zZTbGsaljOa90g+ZlDZJfzmIiPQBXq+H5vp6mupqaaqro6mulsa6Wv/68bamulqa/UMN2gfhrvbEGuPAGRXVrlfV93DFDWxbdx5rj2q37HIREekiLOLEEHss/Ib7A64jLFz/LokEUEiG5VCeOq5sfwFu4yE5PYVD5Ye4ZWRo3b9cRKS/83o8NNXXdRhyGztoa6qro7Guhub6+tMeNzI6hqjYWKJi44iMjsYVn0CEP/QeD7fRx4PwyWHX/6xZFER6l5AMy6Es/0AeAHWpjVAOk9MmB7kiEZG+y+vxnDA1WENNNQ3VvinCjuzZw3tb17ULvb7g29xwmtBrDFHRMUTFxvmCb1w8iekZ/vU4XP4wHBUXR1RM3PHtYmJxhIX13AcXkZChsHyWiouKMMCBsEM4jIPxKeODXZKISK/RFn6rq9oF4Boaa6ra5sVtaJsjt5qmutqOD2QMYRGRtCYnExUTR3R8AsmDh/qDbQfB1x+GI6OjcTgUekWk6xSWz1JJdRkJzji2V25ndNJoop3RwS5JRCRoPK2tbXc/a6g+dne06uPr7W8UUV1FU31dxwcy5oRpwNKGDsOVkNh2w4jo+PbLCUTFxbFmzceaIUFEup3C8lnwtngoc1czKDWd7WXbmTd8XrBLEhEJOE+rm/qqKl/vb3UV9dWVNPjX66sq/W2+9c56fo1xEBXn6/GNjk8g7YLMtqB7bI7c6ITjy1GxserxFZGQpLB8FmoPl1NnmohNiaXOXaf5lUWk12h1u2nwh95jQfdYEPYF4+OBuLPe3wiXi+iERKITkkgZPJSh4yb610/s9VX4FZG+JCTDcqhOHVeQcxSA2qQGKNHFfSISXF6Pxxd2Kyrahd6OeoArO53pIcLlIiYxieiERFKGXsDQCZOJSUhsa4tOSCQm0ffsjIzq4U8oIhJ8IRmWQ3XquKK8QgD2R+QSHxHPsPhhQa5IRPoi6/XSWFtDXWUFdZXl1FVUUF9ZQV1FuW+90rdeX13V4a2DI1zRbWE3begwoidO9oXehCSiE5OIOdYbnJio2wOLiJxBSIblUFVcXkKUI4K1tet1MxIROWvWWpob6v2ht30A9i8fC8ZVFXg9nlP2d8UnEJuUTGxSMgMys4hN9i3HJCYTk5RETEISroQEBWARkQBSWO4i67WUNVaSGpdEbnUu1w+/PtgliUgIcTc1tfX61lVWUH+sF7iiol0YrqC1pfmUfSOjY4hNTiEmKZmh4yf6AnBSSlsYjk1OISYxibBwZxA+mYhI/6aw3EUtZQ1UUs/whCQsVhf3ifQjrW43deVl1JaXUlte5n/4l8t8zx1dqVQXfgAAFi1JREFUFBceEekPvCmkjxxNjL9X2PdIISY5mdjEZJxRGgssIhKqFJa7qHjfUTzGS11cA6bZMDF1YrBLEpEA8LS2Ul9ZQc1J4beuoqwtGDdUV52yX1RsHHEpqcSlppExZpxvOSXVH4h9vcIRrmgN1xIR6eUUlruo8FABALsj95HlyiIuIi7IFYnImXi9Hhqqqk7sCS4vpbbseO9wfVUV1npP2C/CFd0WhAcMz/IH4bS2QByXkqqZIURE+gmF5S4qKinCgYONjZuYkzkn2OWICL4L5hprqqksLKCyqIDKwnwO7NhO4V+W+YJw5akXyoVHRrYF32GTphKXmnpSGE4jMlp35hQRER+F5S4qra0gMSKWanc1k1I1XlmkJzXV11HVLhBXFhZQWVhAVVEBzQ3H5w92hIURHhNLzJChDBk74dQe4dQ0omJiNTRCRES6TGG5Czx1LZR5q0mOSQR0MxKR7uBuaqKyyBeAj4XhysJ8KosKaKypPr6hMcSnppE0aDAXXjGT5EEZJA7KIGnQYOJTB/DxJ58wc+bMoH0OERHpW0IyLIfaHfwqD5TQZNzUx9QT64xlROKIYJck0iu1ut1UFxe1heCqdoG4rqL8hG1jkpJJGpTByOwZJA0aTOKgDJIHDSZhQDrhERFB+gQiItLfhGRYDrU7+BXm5gGwN2I/E1Mn4jCOIFckErq8Hg81pSUnDZnIp6qogJrS0hMupouKiydpUAYXTJhM0qDBJPl7iBPTBxER5QripxAREfEJybAcagoLfLe53uDZzIK0BcEtRiRENNbWUFGQT2VBHhWF/ueCfKqLC/G0trZtF+GKJmlQBukjxzD2ytn+QJxBUvpgomJjg/gJREREzkxhuQtKqsqIdkTR7GjWzUikX/G0uqkqKqKiMI/KgnwqCvzPhfk01da0becICydxYDpJGUPImjbd10ucMZik9AyiExJ1QZ2IiPRaCstnYN1eyloqccX4xkhqJgzpa6y1NFRX+UNw3vHe4oI8qkuKsd7jwyaiExJJzhjCqOmXkjxoMEkZQ0jO8I0jdoSFBfFTiIiIdA+F5TNoyK+imgaiosLJjM8kMSox2CWJnJPWlhbfOOKCvJOGT+SfMP1auDOCxEEZDBg2gjGXXkVyhq+XODljCJHRMUH8BCIiIj1PYfkMivbnYQ3sC8/REAwJadZamhvqqS0rpaaslJqyEt8FdgX5VBTkU1NWAta2bR+bkkrysenXMga39RTHp6ZhHLqIVUREBBSWz6jwqO8217uce7k+bV6Qq5H+zNPqpra83H+7Zl8gri0rpca/XlteSktj4wn7OKNcJA3KIGP0hYy/eo6/l3gIyYMG44zS7ZpFRETORGH5DIrLSggnjDpnnXqWpdtYa2msrfGH4JLjYbi8rC0Q11dVntAzDL4xxHEpqSQNGsywiRcRl5pGfGoacalpxKWkEZOYpIvrREREzoPC8mlYr6WkvoLoiAhcThcjE0PjJinS+7ibm6gtL2vrDa4tLz1hubasjFZ3ywn7hEdEtoXf4UOziUtJbQvC8alpxKak4oyIDNInEhER6R8Ulk+jtaKRCmoxER4mpE4g3KE/Ljk9r9dDZUE+Rbn7KcrdR3FuDlXFhTS2m2YNAGOITUwiLjWNtMwssrIvIS4ljbjUVOJTfIHYFRevXmEREZEgU/o7jbLcItzGQ0H4EU0ZJ6ew1lJTWuwPxv5wfCAXd5Nv3LAzysXA4VmMmnEZ8akDfD3C/iAcm5xMWLgzyJ9AREREzkRh+TQKDvpuc53rOsTdaV8OcjUSbHWVFRTl7qc4d58vHB/IabsxR1h4OGmZIxh/9WzSs0aTnjWKpIzBOByae1hERKQ3U1g+jaKiIgxQHlmhi/v6maa6OooO7KfY32NclLufuopyAIxxkDL0AkZmzyA9axTpWaNJvWCYeopFRET6oJAMy8aY+cD8kSODe0FdaU05UQ4n6fHppLpSg1qLdB93UxPFh3L9wdgXjquKCtteTxqUwZCxE0jPGs3ArJEMzMzStGsiIiL9REiGZWvtUmBpdnb214JVg7fBTVlrNe6oFianTQ5WGRJgnlY3ZUcOt/UWF+Xup/zoEaz13dI5NiWV9BGjmDDzWl84HjGSqNjYIFctIiIiwRKSYTkU1Bwup87RRIWzhBlpVwS7HDkHnlY35XlHKTmYy5FP/sory9+l9PBBPG43AFFx8aRnjWLkxZe0DaeISUwKctUiIiISShSWO1GY47u474grn4fVsxzyWhobKDl8kNJDByg5dICSgwcozzuMp7UVAIczgsGjxjBl7nx/MB5FfNpATc0mIiIip6Ww3ImifN+Y1dLYcsYkjQlyNdJeQ3UVJQdzKfYH49JDuVQWFbbd3c4Vn8CAzBFMvfFWBgwbzoDhWWzds49Zs2cHuXIRERHpbRSWO1FcUUo4YWQNyMIZplkOgsV6vRQfyOHA5g0U5e6j9NAB6ior2l6PTxvIgMwRjL1yFgMysxgwfASxSSmn9BibfTk9XbqIiIj0AQrLHbCtXsqaKsHZqinjgsDd0syR7VvJ3biWAxvXUV9ViTEOkgcPYeiEyQzIHOELxpkjdPGdiIiIdCuF5Q40F9VRaeqpddYwY4Au7usJ9VWV5G5cx4FN6zi8bQutLc1EuFxkTppKVvYMhk/JxhUXH+wyRUREpJ9RWO5AcU4+HuOlMKpIt7nuJtZayo4cInfjOnI3rqUoZx8AcalpTJh1LVnTpjNk3ETCnRoCIyIiIsGjsNyBwsP5ANQkNzAwZmCQq+k7PK1uju7aQe6GtRzYtI6a0hIA0keO5vIvfpms7BmkXpCpGSpEREQkZCgsd6C4pBiDYeSQ4N5BsC+oKS3h0NZNHNyykSM7ttDS2Eh4RCQXTJzMjNvuYsTUi4lNSg52mSIiIiIdUlg+ibWWkvoKwgxMGqghGGfL3dJM3q4dHNq6iUNbNlJR4JuvOi4ljTGXXUXWtOlcMGEyzkjdLlpERERCn8LySTzVzVTYWpoiGnSb6y6w1lKRn8ehrRs5tHUTebt20OpuIczpZMjYCUy6Zi6Zk6eRPHiIhleIiIhIr6OwfJLK3BIaTQvlURWMTRkb7HJCUnNDPUe2b+WgPyDXlpUCkJwxhEnX3kDm5KkMGTtevcciIiLS6yksn6TwoG/YgB1oiAyLDHI1oaM8/yg56z/n4OYNFOzbjfV6iXC5uGDCRVxy211kTp5KfNqAYJcpIiIiElAKyycpj24AIDNzRJArCS5rLcUHcshZ/xn7131GRf5RAAZkZjH9ljvJnDyVQaMuJCxcP0IiIiLSdynpnMQMDmdPwh4eyHgg2KX0OK/HQ97uneSs/4yc9Z9TW16KcTgYOm4CF103j5EXX0pcSmqwyxQRERHpMQrLJymKLGJn8s5+c5trd0szh7dtIWfdZ+RuWkdTbQ3hzgiGTZ7CZV+8l6xp03XnPBEREem3FJZPEuOM4bKMy8iIyQh2Kd2muaGeA5vWk7PuMw5u2Yi7uYnI6BhGTL2YkdMvZfjkaTijdHGeiIiIiMLySeZnzWd+1vxglxFQrS0tFOXuI3/PLo7u2s7RndvxelqJSUxi7JUzGXXxpQydMImwcN1aWkRERKS9HgvLxpixwLeAVGCltfb/9dR79zetbjdHdmwhb/dO8vfsojh3H57WVgCSBw9l6rybGXnxpWSMGoNxOIJcrYiIiEjo6lJYNsa8DNwElFhrJ7Rrnwv8JxAG/M5a+3Rnx7DW7gYeNcY4gD8ACssBVlVUyLaVH7Bj1XIaa2twhIUzMGskU264mcEXjidj9IVExycEu0wRERGRXqOrPcuLgF/jC7kAGGPCgBeBa4E8YL0x5l18wfkXJ+3/kLW2xBhzM/B3wB/Ps27x87S2cmDjOraueJ/D2zZjHA6yps1g0pzrGTJugm4MIiIiInIejLW2axsakwm8d6xn2RhzKfCktfZ6//oPAay1Jwfljo71Z2vtjZ289jDwMMDAgQOnLV68uEv1nYu6ujpiY2O77fjdqbm2mrLd2ynbvZ3WhnqcMXGkjZtEytiJRMT0zs/UnXrzuZazo3Pdf+hc9x861/1HsM71rFmzNlprszt67XzGLA8GjrZbzwNmdLaxMWYmcDsQCSzrbDtr7UJgIUB2dradOXPmeZR4eqtXr6Y7jx9oXq+Hg5s3sG3FBxzYvAGAEVOymXTNXIZflI0jLCzIFYau3nau5dzpXPcfOtf9h851/xGK57rHLvCz1q4GVvfU+/Ul9VWVbFv5AdtXfkRteSkxiUlcctsXmTj7et1iWkRERKQbnU9YzgeGtlsf4m+TAKkoyGfDe2+y668r8bS2MmzSFGY+8FWyps3QbaZFREREesD5JK71wChjzHB8Iflu4EuBKMoYMx+YP3LkyEAcrtcp3L+X9e8uYf/6zwgLD2fCrGuZOu9WkjMGB7s0ERERkX6lq1PHvQrMBFKNMXnAP1trf2+MeQz4EN8MGC9ba3cGoihr7VJgaXZ29tcCcbzewFrLwS0bWP/uEvJ27SAyJoYZt36RKXNvIiYxKdjliYiIiPRLXQrL1tp7Omlfxmku1pMz87S2svdva1j/7hLKjh4mNiWVmfd/lYmzryPCFR3s8kRERET6NQ18DZKWpka2r/yIjX9+m9ryUlKGXMDcr3+HCy+/SredFhEREQkRIRmW+/KY5ZqyErat+ICtHy2jqb6OIWMncM1Xv87wKdkYY4JdnoiIiIi0E5Jhua+NWbZeL4e3bWbL8mUc2Lgei2Vk9gwuvvlOMkZfGOzyRERERKQTIRmW+4rG2hp2rF7BtuXvU1VciCs+gem33smkOXM1P7KIiIhIL6CwHGDWWgr372Xr8mXs/exjPG43gy8cz2V3fZlR0y8j3KnxyCIiIiK9hcJygLibmtj96Wq2fvQ+JYdycUa5mDDrOiZfewNpF2QGuzwREREROQchGZZ70wV+TXV1fPbG/7Fj9QpaGhtIvSCTa776dcZeMVNTv4mIiIj0ciEZlnvLBX55u3aw7Nf/Tn1VBaMvuYLJ181j8JhxmtVCREREpI8IybAc6jytrXz2xqusfft1Egemc89P/430kaODXZaIiIiIBJjC8lmqKipk2a+epTBnLxNmXcusBQ8TEeUKdlkiIiIi0g0UlrvIWsuuNX9h5csv4QhzcNO3n2DMpVcEuywRERER6UYhGZZD7QK/pvo6VvzuN+z92xqGjJ3ADY99l/hUzZMsIiIi0teFZFgOpQv88vbs5P1f/zu15WVcftd9TL/1ThyOsGCXJSIiIiI9ICTDcijwejx8tmQxa998jfgBA7jnp//GoFFjgl2WiIiIiPQgheUOVJcU8edfPUvhvj2Mv3oOsx98RHMmi4iIiPRDCssn2fvZJ3z02xcwxnDjN/+BCy+/OtgliYiIiEiQKCyfxFovacMymffY94hP00V8IiIiIv1ZSIblYM6GceFlVzHmkiswDkePv7eIiIiIhJaQTITW2qXW2ocTEhKC8v4KyiIiIiICIRqWRURERERCgcKyiIiIiEgnFJZFRERERDqhsCwiIiIi0gmFZRERERGRTigsi4iIiIh0IiTDsjFmvjFmYXV1dbBLEREREZF+LCTDcrDnWRYRERERgRANyyIiIiIioUBhWURERESkEwrLIiIiIiKdUFgWEREREemEwrKIiIiISCcUlkVEREREOqGwLCIiIiLSiZAMy7opiYiIiIiEgpAMy7opiYiIiIiEAmOtDXYNnTLGlAKHu/EtUoGybjy+hA6d6/5D57r/0LnuP3Su+49gneth1tq0jl4I6bDc3YwxG6y12cGuQ7qfznX/oXPdf+hc9x861/1HKJ7rkByGISIiIiISChSWRUREREQ60d/D8sJgFyA9Rue6/9C57j90rvsPnev+I+TOdb8esywiIiIicjr9vWdZRERERKRTCssiIiIiIp3ot2HZGDPXGLPXGJNjjHki2PXI2THGDDXGrDLG7DLG7DTGfMvfnmyMWW6M2e9/TvK3G2PMC/7zvc0YM7XdsR7wb7/fGPNAsD6TnJ4xJswYs9kY855/fbgxZq3/nL5mjInwt0f613P8r2e2O8YP/e17jTHXB+eTyOkYYxKNMW8YY/YYY3YbYy7V97pvMsZ8x//39w5jzKvGmCh9r/sGY8zLxpgSY8yOdm0B+x4bY6YZY7b793nBGGO69QNZa/vdAwgDcoERQASwFRgX7Lr0OKtzOAiY6l+OA/YB44BngCf87U8Av/QvzwPeBwxwCbDW354MHPA/J/mXk4L9+fTo8Jx/F/g/4D3/+uvA3f7ll4C/8y9/HXjJv3w38Jp/eZz/ux4JDPf/HRAW7M+lxynn+X+Ar/qXI4BEfa/73gMYDBwEXP7114EF+l73jQdwFTAV2NGuLWDfY2Cdf1vj3/eG7vw8/bVneTqQY609YK1tARYDtwS5JjkL1tpCa+0m/3ItsBvfX7634PvHFv/zrf7lW4A/WJ/PgURjzCDgemC5tbbCWlsJLAfm9uBHkS4wxgwBbgR+5183wGzgDf8mJ5/rYz8DbwBz/NvfAiy21jZbaw8COfj+LpAQYYxJwPeP7O8BrLUt1toq9L3uq8IBlzEmHIgGCtH3uk+w1q4BKk5qDsj32P9avLX2c+tLzn9od6xu0V/D8mDgaLv1PH+b9EL+X8dNAdYCA621hf6XioCB/uXOzrl+FnqH54HvA17/egpQZa1t9a+3P29t59T/erV/e53r0DccKAX+2z/k5nfGmBj0ve5zrLX5wLPAEXwhuRrYiL7XfVmgvseD/csnt3eb/hqWpY8wxsQCS4BvW2tr2r/m/x+n5kbs5YwxNwEl1tqNwa5Ful04vl/d/j9r7RSgHt+va9voe903+Mer3oLvP0gZQAzq/e83etv3uL+G5XxgaLv1If426UWMMU58QfkVa+2b/uZi/69o8D+X+Ns7O+f6WQh9lwM3G2MO4RsyNRv4T3y/qgv3b9P+vLWdU//rCUA5Ote9QR6QZ61d619/A1941ve677kGOGitLbXWuoE38X3X9b3uuwL1Pc73L5/c3m36a1heD4zyX3Ubge9igXeDXJOcBf9Ytd8Du621z7V76V3g2BWzDwDvtGu/33/V7SVAtf/XQR8C1xljkvw9Hdf52yREWGt/aK0dYq3NxPdd/Yu19l5gFXCnf7OTz/Wxn4E7/dtbf/vd/qvqhwOj8F0kIiHCWlsEHDXGjPE3zQF2oe91X3QEuMQYE+3/+/zYudb3uu8KyPfY/1qNMeYS/8/O/e2O1T168urIUHrgu/pyH74rZ38c7Hr0OOvzdwW+X+FsA7b4H/PwjWFbCewHVgDJ/u0N8KL/fG8Hstsd6yF8F4XkAA8G+7PpcdrzPpPjs2GMwPePYg7wJyDS3x7lX8/xvz6i3f4/9v8M7KWbr57W45zP8UXABv93+218V8Hre90HH8C/AHuAHcAf8c1ooe91H3gAr+Ibi+7G9xujrwTyewxk+39ucoFf478jdXc9dLtrEREREZFO9NdhGCIiIiIiZ6SwLCIiIiLSCYVlEREREZFOKCyLiIiIiHRCYVlEREREpBMKyyIiIiIinVBYFhERERHpxP8HkJVCDcA8+sEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(1, 1, figsize=(12, 8))\n",
    "\n",
    "for name in names:\n",
    "    ax.plot(df_mean['Num'], df_mean[name], label=name)\n",
    "    \n",
    "plt.yscale('log')\n",
    "ax.legend()\n",
    "ax.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('result/result_time_num.csv', index_label=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## EXTRA EXPERIMENTS"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dependacne of Num points"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "DE = DimEst()\n",
    "DG = DataGenerator()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['Helix1d', 'Helix2d', 'Helicoid', 'Spiral', 'Roll', 'Scurve', 'Star', 'Moebius', 'Sphere', 'Norm', 'Uniform', 'Cubic', 'Affine_3to5', 'Affine', 'Nonlinear_4to6', 'Nonlinear', 'Paraboloid'])"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DG.dict_gen.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "steps = [\n",
    "    ('Sphere',  5,  3),\n",
    "    ('Sphere',  10, 5),\n",
    "    ('Sphere',  15, 10),\n",
    "    ('Sphere',  35, 30),\n",
    "    ('Sphere',  55, 50),\n",
    "    ('Sphere',  75, 70),\n",
    "    ('Sphere',  100, 75),\n",
    "    ('Sphere',  125, 100),\n",
    "    \n",
    "    ('Affine_3to5',  5,  3),\n",
    "    ('Affine',       5,  5),\n",
    "    ('Affine',       10, 10),\n",
    "    ('Affine',       20, 20),\n",
    "    \n",
    "    ('Nonlinear_4to6',  6,  4),\n",
    "    ('Nonlinear',       8,  4),\n",
    "    ('Nonlinear',       36, 6),\n",
    "    ('Nonlinear',       64, 8),\n",
    "    ('Nonlinear',       72, 12),\n",
    "    \n",
    "    ('Spiral',   3,  1),\n",
    "    ('Spiral',   13, 1),\n",
    "    ('Helix1d',  3,  1),\n",
    "    ('Helix1d',  13, 1),\n",
    "    ('Helix2d',  3,  2),\n",
    "    ('Helix2d',  13, 2),\n",
    "    ('Helicoid', 3,  2),\n",
    "    ('Helicoid', 13, 2),\n",
    "    ('Roll',     3,  2),\n",
    "    ('Roll',     13, 2),\n",
    "    ('Scurve',   3,  2),\n",
    "    ('Scurve',   13, 2),\n",
    "    ('Moebius',  3,  2),\n",
    "    \n",
    "    ('Uniform', 10, 5),\n",
    "    ('Uniform', 15, 10),\n",
    "    ('Uniform', 35, 30),\n",
    "    ('Uniform', 55, 50),\n",
    "    ('Uniform', 75, 70),\n",
    "     \n",
    "    ('Norm', 20, 20),\n",
    "    ('Norm', 50, 20),\n",
    "    ('Norm', 50, 50),\n",
    "    ('Norm', 70, 50),\n",
    "    \n",
    "    ('Paraboloid', 12, 3),\n",
    "    ('Paraboloid', 21, 6),\n",
    "    ('Paraboloid', 30, 9),\n",
    "    ('Paraboloid', 39, 12),\n",
    "    ('Paraboloid', 48, 15),\n",
    "    \n",
    "    ('Cubic', 5,  3),\n",
    "    ('Cubic', 35, 30),\n",
    "    ('Cubic', 90, 70),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "47"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "N_steps = np.concatenate((np.arange(100, 1000, 200), \n",
    "                          np.arange(1000, 2000, 250),\n",
    "                          np.arange(2000, 2501, 500)))                      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 100,  300,  500,  700,  900, 1000, 1250, 1500, 1750, 2000, 2500])"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "N_steps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_steps = [(N, name, D, d) for name, D, d in steps for N in N_steps]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "517"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(all_steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_args(const_args, step):\n",
    "    args = {}\n",
    "    args['noise'] = const_args['noise']\n",
    "    args['type_sample'] = const_args['type_sample']\n",
    "    args['N'] = step[0]\n",
    "    args['name'] = step[1]\n",
    "    args['D'] = step[2]\n",
    "    args['d'] = step[3]\n",
    "    return args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  5%|▌         | 133/2585 [04:50<54:49,  1.34s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Affine', 'D': 20, 'd': 20}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  6%|▌         | 155/2585 [05:31<1:26:18,  2.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Nonlinear', 'D': 8, 'd': 4}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 12%|█▏        | 298/2585 [09:25<39:56,  1.05s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Roll', 'D': 13, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 14%|█▍        | 364/2585 [10:14<35:10,  1.05it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Uniform', 'D': 35, 'd': 30}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 17%|█▋        | 430/2585 [12:59<1:56:00,  3.23s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Norm', 'D': 70, 'd': 50}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 18%|█▊        | 463/2585 [14:30<1:31:24,  2.58s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Paraboloid', 'D': 30, 'd': 9}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 18%|█▊        | 474/2585 [14:49<1:27:14,  2.48s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Paraboloid', 'D': 39, 'd': 12}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 19%|█▉        | 496/2585 [15:37<1:07:26,  1.94s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Cubic', 'D': 5, 'd': 3}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|█▉        | 507/2585 [15:57<1:19:22,  2.29s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Cubic', 'D': 35, 'd': 30}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 21%|██        | 530/2585 [16:51<1:41:33,  2.97s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Sphere', 'D': 5, 'd': 3}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 23%|██▎       | 595/2585 [19:10<1:26:23,  2.60s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Sphere', 'D': 100, 'd': 75}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 25%|██▌       | 650/2585 [22:52<51:24,  1.59s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Affine', 'D': 20, 'd': 20}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 28%|██▊       | 716/2585 [25:40<2:30:51,  4.84s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Spiral', 'D': 3, 'd': 1}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 29%|██▉       | 749/2585 [26:32<39:37,  1.29s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Helix1d', 'D': 13, 'd': 1}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 36%|███▌      | 925/2585 [30:15<1:37:50,  3.54s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Norm', 'D': 50, 'd': 20}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 39%|███▉      | 1013/2585 [33:59<56:06,  2.14s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Cubic', 'D': 5, 'd': 3}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 49%|████▉     | 1277/2585 [45:17<26:39,  1.22s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Helix2d', 'D': 3, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 55%|█████▌    | 1430/2585 [48:13<1:45:58,  5.51s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Norm', 'D': 20, 'd': 20}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 59%|█████▉    | 1519/2585 [52:06<1:04:08,  3.61s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Paraboloid', 'D': 48, 'd': 15}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 62%|██████▏   | 1596/2585 [54:34<28:35,  1.73s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Sphere', 'D': 35, 'd': 30}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 66%|██████▌   | 1706/2585 [1:00:37<32:42,  2.23s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Nonlinear', 'D': 8, 'd': 4}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 69%|██████▉   | 1783/2585 [1:03:35<22:17,  1.67s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Helix1d', 'D': 13, 'd': 1}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|███████   | 1816/2585 [1:04:10<09:18,  1.38it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Helicoid', 'D': 3, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 72%|███████▏  | 1871/2585 [1:05:00<11:00,  1.08it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Scurve', 'D': 13, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 77%|███████▋  | 1992/2585 [1:09:08<42:39,  4.32s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Paraboloid', 'D': 12, 'd': 3}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|███████▊  | 2025/2585 [1:10:15<23:05,  2.47s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Paraboloid', 'D': 39, 'd': 12}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 83%|████████▎ | 2157/2585 [1:15:56<55:59,  7.85s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Sphere', 'D': 125, 'd': 100}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 86%|████████▌ | 2212/2585 [1:18:46<12:09,  1.96s/it]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Nonlinear_4to6', 'D': 6, 'd': 4}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 87%|████████▋ | 2245/2585 [1:19:40<10:57,  1.93s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Nonlinear', 'D': 64, 'd': 8}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 89%|████████▉ | 2311/2585 [1:22:26<05:48,  1.27s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Helix2d', 'D': 3, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 92%|█████████▏| 2366/2585 [1:23:17<03:04,  1.18it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Roll', 'D': 13, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 92%|█████████▏| 2376/2585 [1:23:26<03:01,  1.15it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Scurve', 'D': 3, 'd': 2}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 97%|█████████▋| 2520/2585 [1:28:18<03:41,  3.40s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Paraboloid', 'D': 21, 'd': 6}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 99%|█████████▉| 2553/2585 [1:29:20<01:38,  3.08s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WRONG with {'noise': 0, 'type_sample': 'uniform', 'N': 100, 'name': 'Paraboloid', 'D': 48, 'd': 15}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2585/2585 [1:30:17<00:00,  1.60s/it]\n"
     ]
    }
   ],
   "source": [
    "names = DE.names\n",
    "n_repetition = 5\n",
    "\n",
    "mpe = []\n",
    "args = {'noise': 0, 'type_sample': 'uniform'}\n",
    "df = run_test(run_one_test, all_steps, args, n_repetition)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "df, _ = df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "dfr = df.round()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['MLE', 'GeoMLE', 'MIND', 'DANCo', 'ESS', 'PCA', 'CD', 'Hein']"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "dfr[names] = abs(dfr[names].sub(dfr['RealDim'], axis=0)).mul(dfr['RealDim']**-1, axis=0)\n",
    "\n",
    "mpe = dfr.groupby('Num')[names].mean().values\n",
    "std = dfr.groupby('Num')[names].std().values / dfr.groupby('Num')[names].count().values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(1, 1, figsize=(12, 8))\n",
    "\n",
    "for mpe_i, std_i, name in zip(mpe.T, std.T, names):\n",
    "    ax.plot(N_steps, mpe_i, label=name)\n",
    "    ax.fill_between(N_steps, mpe_i-std_i, mpe_i+std_i, alpha=0.4)\n",
    "    \n",
    "ax.legend(loc=5)\n",
    "ax.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('result/result_num.csv', index_label=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}