01 - cleaning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "#plotting\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from imblearn.over_sampling import ADASYN\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from IPython.display import Audio\n",
    "\n",
    "# create a PING noise to let you know modeling is done\n",
    "sound_file = 'glass_ping-Go445-1207030150.mp3'\n",
    "play_sound = Audio(sound_file, autoplay=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df = pd.read_csv('Fraud/train_transaction.csv')\n",
    "\n",
    "# test is based on Kaggle competition, and does not include targets\n",
    "test_df = pd.read_csv('Fraud/test_transaction.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_df['TransactionDT'].plot(kind='hist', bins=20, figsize=(15,4))\n",
    "test_df['TransactionDT'].plot(kind='hist', bins=20, figsize=(15,4));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total transaction amount: $79738948.74\n",
      "Total fraud transaction amount: $3083844.86\n",
      "Fraud represents 3.87% of total.\n"
     ]
    }
   ],
   "source": [
    "total_trans = train_df['TransactionAmt'].sum()\n",
    "\n",
    "fraud_trans = train_df.query(\"isFraud == 1\")['TransactionAmt'].sum()\n",
    "\n",
    "print(f'Total transaction amount: ${round(total_trans,2)}')\n",
    "print(f'Total fraud transaction amount: ${round(fraud_trans,2)}')\n",
    "print(f'Fraud represents {round((fraud_trans/total_trans)*100,2)}% of total.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "## un-comment if you'd like to reduce the memory usage - \n",
    "## potentially breaks some functions later\n",
    "\n",
    "# def reduce_mem_usage(df, verbose=True):\n",
    "#     numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']\n",
    "#     start_mem = df.memory_usage().sum() / 1024**2    \n",
    "#     for col in df.columns:\n",
    "#         col_type = df[col].dtypes\n",
    "#         if col_type in numerics:\n",
    "#             c_min = df[col].min()\n",
    "#             c_max = df[col].max()\n",
    "#             if str(col_type)[:3] == 'int':\n",
    "#                 if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:\n",
    "#                     df[col] = df[col].astype(np.int8)\n",
    "#                 elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:\n",
    "#                     df[col] = df[col].astype(np.int16)\n",
    "#                 elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:\n",
    "#                     df[col] = df[col].astype(np.int32)\n",
    "#                 elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:\n",
    "#                     df[col] = df[col].astype(np.int64)  \n",
    "#             else:\n",
    "#                 if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:\n",
    "#                     df[col] = df[col].astype(np.float16)\n",
    "#                 elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:\n",
    "#                     df[col] = df[col].astype(np.float32)\n",
    "#                 else:\n",
    "#                     df[col] = df[col].astype(np.float64)    \n",
    "#     end_mem = df.memory_usage().sum() / 1024**2\n",
    "#     if verbose: print('Mem. usage decreased to {:5.2f} Mb ({:.1f}% reduction)'.format(end_mem, 100 * (start_mem - end_mem) / start_mem))\n",
    "#     return df\n",
    "\n",
    "\n",
    "# train_df = reduce_mem_usage(train_df)\n",
    "# test_df = reduce_mem_usage(test_df)\n",
    "\n",
    "# play_sound"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test_df does not contain targets, and only used for Kaggle competition\n",
    "del test_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df = train_df.iloc[:, :32].drop(columns=['dist1','dist2','R_emaildomain'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>TransactionID</th>\n",
       "      <th>isFraud</th>\n",
       "      <th>TransactionDT</th>\n",
       "      <th>TransactionAmt</th>\n",
       "      <th>card1</th>\n",
       "      <th>card2</th>\n",
       "      <th>card3</th>\n",
       "      <th>card5</th>\n",
       "      <th>addr1</th>\n",
       "      <th>addr2</th>\n",
       "      <th>...</th>\n",
       "      <th>P_emaildomain_web.de</th>\n",
       "      <th>P_emaildomain_windstream.net</th>\n",
       "      <th>P_emaildomain_yahoo.co.jp</th>\n",
       "      <th>P_emaildomain_yahoo.co.uk</th>\n",
       "      <th>P_emaildomain_yahoo.com</th>\n",
       "      <th>P_emaildomain_yahoo.com.mx</th>\n",
       "      <th>P_emaildomain_yahoo.de</th>\n",
       "      <th>P_emaildomain_yahoo.es</th>\n",
       "      <th>P_emaildomain_yahoo.fr</th>\n",
       "      <th>P_emaildomain_ymail.com</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>355303</td>\n",
       "      <td>3342303</td>\n",
       "      <td>0</td>\n",
       "      <td>8796239</td>\n",
       "      <td>171.0</td>\n",
       "      <td>2157</td>\n",
       "      <td>111.0</td>\n",
       "      <td>150.0</td>\n",
       "      <td>226.0</td>\n",
       "      <td>494.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>194819</td>\n",
       "      <td>3181819</td>\n",
       "      <td>0</td>\n",
       "      <td>4388390</td>\n",
       "      <td>272.0</td>\n",
       "      <td>16297</td>\n",
       "      <td>544.0</td>\n",
       "      <td>150.0</td>\n",
       "      <td>117.0</td>\n",
       "      <td>123.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 93 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        TransactionID  isFraud  TransactionDT  TransactionAmt  card1  card2  \\\n",
       "355303        3342303        0        8796239           171.0   2157  111.0   \n",
       "194819        3181819        0        4388390           272.0  16297  544.0   \n",
       "\n",
       "        card3  card5  addr1  addr2  ...  P_emaildomain_web.de  \\\n",
       "355303  150.0  226.0  494.0   87.0  ...                     0   \n",
       "194819  150.0  117.0  123.0   87.0  ...                     0   \n",
       "\n",
       "        P_emaildomain_windstream.net  P_emaildomain_yahoo.co.jp  \\\n",
       "355303                             0                          0   \n",
       "194819                             0                          0   \n",
       "\n",
       "        P_emaildomain_yahoo.co.uk  P_emaildomain_yahoo.com  \\\n",
       "355303                          0                        1   \n",
       "194819                          0                        1   \n",
       "\n",
       "        P_emaildomain_yahoo.com.mx  P_emaildomain_yahoo.de  \\\n",
       "355303                           0                       0   \n",
       "194819                           0                       0   \n",
       "\n",
       "        P_emaildomain_yahoo.es  P_emaildomain_yahoo.fr  \\\n",
       "355303                       0                       0   \n",
       "194819                       0                       0   \n",
       "\n",
       "        P_emaildomain_ymail.com  \n",
       "355303                        0  \n",
       "194819                        0  \n",
       "\n",
       "[2 rows x 93 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# find catagorical columns\n",
    "cat_cal = [x for x in train_df if train_df[x].dtype == 'O']\n",
    "\n",
    "# dummy features\n",
    "dummy_df = pd.get_dummies(train_df, columns=cat_cal, drop_first=True)\n",
    "\n",
    "dummy_df.sample(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting Features and Target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = dummy_df.drop(columns=['isFraud'])\n",
    "y = dummy_df['isFraud']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Replacing NA Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TransactionID      0.000000\n",
       "isFraud            0.000000\n",
       "TransactionDT      0.000000\n",
       "TransactionAmt     0.000000\n",
       "ProductCD          0.000000\n",
       "card1              0.000000\n",
       "card2              1.512683\n",
       "card3              0.265012\n",
       "card4              0.267044\n",
       "card5              0.721204\n",
       "card6              0.266028\n",
       "addr1             11.126427\n",
       "addr2             11.126427\n",
       "P_emaildomain     15.994852\n",
       "C1                 0.000000\n",
       "C2                 0.000000\n",
       "C3                 0.000000\n",
       "C4                 0.000000\n",
       "C5                 0.000000\n",
       "C6                 0.000000\n",
       "C7                 0.000000\n",
       "C8                 0.000000\n",
       "C9                 0.000000\n",
       "C10                0.000000\n",
       "C11                0.000000\n",
       "C12                0.000000\n",
       "C13                0.000000\n",
       "C14                0.000000\n",
       "D1                 0.214888\n",
       "dtype: float64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show % of NAs in columns\n",
    "(train_df.isna().sum() / train_df.shape[0]) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3min 4s, sys: 1.9 s, total: 3min 6s\n",
      "Wall time: 3min 5s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "167282"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "# find the transactions that have at least one na\n",
    "na_lst = []\n",
    "for i in train_df.index:\n",
    "    if train_df.iloc[i].isnull().any() > 0:\n",
    "        na_lst.append(i)\n",
    "        \n",
    "len(na_lst)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Because there are so many observations with NAs, we're going to impute those values, instead of dropping them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# lots of outliers\n",
    "train_df.iloc[: ,3:].drop(columns=['card1']).plot(kind='box', figsize=(20,8));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "na_cols = X_train.columns[X_train.isna().sum()>0]\n",
    "\n",
    "test_na_cols = X_test.columns[X_test.isna().sum()>0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "card2 mean 362.44545010625933\n",
      "card2 median 361.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "card3 mean 153.1657112973553\n",
      "card3 median 150.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "card5 mean 199.17866353724602\n",
      "card5 median 226.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "addr1 mean 290.51643775124865\n",
      "addr1 median 299.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "addr2 mean 86.8081830917286\n",
      "addr2 median 87.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "D1 mean 94.72293899269053\n",
      "D1 median 3.0\n"
     ]
    },
    {
     "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": [
    "# showing the best values of dataset to replace NAs with\n",
    "\n",
    "for col in X_test[na_cols]:\n",
    "    if X_test[col].dtype != 'O':\n",
    "        X_test[col].plot(kind='hist', label='')\n",
    "        plt.title(col)\n",
    "        ymin, ymax = plt.ylim()\n",
    "        plt.vlines(X_test[col].mean(), ymin, ymax, color='blue', label='mean')\n",
    "        plt.vlines(X_test[col].median(), ymin, ymax, color='red', label='median')\n",
    "        print(f'{col} mean {X_test[col].mean()}')\n",
    "        print(f'{col} median {X_test[col].median()}')\n",
    "        plt.legend()\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/envs/learn-env/lib/python3.6/site-packages/pandas/core/generic.py:6287: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self._update_inplace(new_data)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fill NAs with median values\n",
    "for column in X_train[na_cols]:\n",
    "    median = X_train[column].median()\n",
    "    X_train[column].fillna(median, inplace=True)\n",
    "    \n",
    "\n",
    "for column in X_train[test_na_cols]:\n",
    "    median = X_test[column].median()\n",
    "    X_test[column].fillna(median, inplace=True)\n",
    "\n",
    "X_test.isna().sum().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ADASYN Over Sampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/envs/learn-env/lib/python3.6/site-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function safe_indexing is deprecated; safe_indexing is deprecated in version 0.22 and will be removed in version 0.24.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((857945, 92), (857945,))"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ada = ADASYN('minority', random_state=42, n_neighbors = 6, n_jobs=-1)\n",
    "\n",
    "columns= X_train.columns\n",
    "X_train, y_train = ada.fit_resample(X_train, y_train)\n",
    "\n",
    "y_train = pd.Series(y_train)\n",
    "X_train = pd.DataFrame(X_train, columns=columns)\n",
    "X_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train.to_pickle('X_train')\n",
    "y_train.to_pickle('y_train')\n",
    "X_test.to_pickle('X_test')\n",
    "y_test.to_pickle('y_test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "learn-env",
   "language": "python",
   "name": "learn-env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}