Sklearn - prediction pipeline v2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "##################################################################################\n",
    "## This uses the model created in sklearn v1, to predict the mask for the \n",
    "## given image\n",
    "##################################################################################\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os \n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "np.random.seed(1337)  # for reproducibility\n",
    "\n",
    "from sklearn import svm, metrics \n",
    "from sklearn.externals import joblib\n",
    "\n",
    "import scipy.misc\n",
    "from scipy.misc import imread\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## Load the model\n",
    "#lin_clf = joblib.load('linearSVC_v1.pkl')\n",
    "lin_clf = joblib.load('grid_linearSVC_v0.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['017532875DDF.jpg',\n",
       " '029E137BB177.jpg',\n",
       " '029E137BB179.jpg',\n",
       " '072CBBB64F88.jpg',\n",
       " '072CBBB64F89.jpg',\n",
       " '0BC2C60F3BA0.jpg',\n",
       " '1468054105.jpg',\n",
       " '1468054163.jpg',\n",
       " '1468054267.jpg',\n",
       " '1468054306.jpg',\n",
       " '1468061314.jpg',\n",
       " '1468061317.jpg',\n",
       " '1468061347.jpg',\n",
       " '1468061368.jpg',\n",
       " '1468061411.jpg',\n",
       " '16B1C9836EB0.jpg',\n",
       " '264882623008.jpg',\n",
       " '2D6D41F5B4A1.jpg',\n",
       " '308A1C309EF5.jpg',\n",
       " '31A25DC60455.jpg',\n",
       " '31A25DC60456.jpg',\n",
       " '34DBA85D4F7A.jpg',\n",
       " '34DBA85D4F7E.jpg',\n",
       " '469CE8C373A0.jpg',\n",
       " '469CE8C373AC.jpg',\n",
       " '5274941329CC.jpg',\n",
       " '5274941329CF.jpg',\n",
       " '52B380E6A3F1.jpg',\n",
       " '52B380E6A3F2.jpg',\n",
       " '549C148B2D40.jpg',\n",
       " '549C148B2D4E.jpg',\n",
       " '558DF2AD8A5F.jpg',\n",
       " '5ABD740F7441.jpg',\n",
       " '5ABD740F744D.jpg',\n",
       " '5C75353DF081.jpg',\n",
       " '5C75353DF082.jpg',\n",
       " '682E28C9537E.jpg',\n",
       " '7D00B08A7B2D.jpg',\n",
       " '85639DF5DF72.jpg',\n",
       " '85639DF5DF7C.jpg',\n",
       " '8A87C3794B59.jpg',\n",
       " 'BB6EB0F294D8.jpg',\n",
       " 'BB6EB0F294DA.jpg',\n",
       " 'C11E6107F913.jpg',\n",
       " 'C3AC084EA711.jpg',\n",
       " 'C4D8C497AB5C.jpg',\n",
       " 'CD406B9D61C4.jpg',\n",
       " 'CD406B9D61CE.jpg',\n",
       " 'D0F6DE661D63.jpg',\n",
       " 'D0F6DE661D64.jpg',\n",
       " 'D0F6DE661D65.jpg',\n",
       " 'D28CDF85BDA3.jpg',\n",
       " 'D91DD06F9B3D.jpg',\n",
       " 'D91DD06F9B3E.jpg',\n",
       " 'DFC9150DB405.jpg',\n",
       " 'E04FFD273493.jpg',\n",
       " 'E04FFD273499.jpg',\n",
       " 'E3084BAFECCD.jpg',\n",
       " 'E89EFB01CEB9.jpg',\n",
       " 'EB2EFD28E53E.jpg',\n",
       " 'F1BFEA74B33D.jpg']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Read the test files\n",
    "ip_files = []\n",
    "test_files = os.listdir('E:\\\\SigTuple_Hack\\\\Test_Data\\\\')\n",
    "ip_files += [f for f in test_files if f.endswith('.jpg') and 'mask' not in f]\n",
    "\n",
    "#ip_files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "E:\\SigTuple_Hack\\Test_Data\\017532875DDF.jpg\n"
     ]
    }
   ],
   "source": [
    "img_file = ip_files[0]\n",
    "input_file = 'E:\\\\SigTuple_Hack\\\\Test_Data\\\\' + img_file\n",
    "mask_file = 'E:\\\\SigTuple_Hack\\\\Test_Data\\\\Mask\\\\' + img_file[:-4] + '-mask.jpg'\n",
    "print input_file\n",
    "    \n",
    "# Load the image & flatten 3d to 2d .\n",
    "imgcolor = imread(input_file)/255.0\n",
    "imgshape = imgcolor.shape\n",
    "imgdf = pd.DataFrame(imgcolor.transpose(2,0,1).reshape(3,-1).transpose(1,0))\n",
    "imgdf.columns = ['Red', 'Green', 'Blue']\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Create mask, using LinearSVC model\n",
    "mask = lin_clf.predict(imgdf)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "    \n",
    "imgmask = mask.reshape(imgshape[0], imgshape[1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "    \n",
    "# Save file\n",
    "scipy.misc.imsave(mask_file, imgmask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "E:\\SigTuple_Hack\\Test_Data\\017532875DDF.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\029E137BB177.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\029E137BB179.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\072CBBB64F88.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\072CBBB64F89.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\0BC2C60F3BA0.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468054105.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468054163.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468054267.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468054306.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468061314.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468061317.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468061347.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468061368.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\1468061411.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\16B1C9836EB0.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\264882623008.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\2D6D41F5B4A1.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\308A1C309EF5.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\31A25DC60455.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\31A25DC60456.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\34DBA85D4F7A.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\34DBA85D4F7E.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\469CE8C373A0.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\469CE8C373AC.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\5274941329CC.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\5274941329CF.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\52B380E6A3F1.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\52B380E6A3F2.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\549C148B2D40.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\549C148B2D4E.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\558DF2AD8A5F.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\5ABD740F7441.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\5ABD740F744D.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\5C75353DF081.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\5C75353DF082.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\682E28C9537E.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\7D00B08A7B2D.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\85639DF5DF72.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\85639DF5DF7C.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\8A87C3794B59.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\BB6EB0F294D8.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\BB6EB0F294DA.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\C11E6107F913.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\C3AC084EA711.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\C4D8C497AB5C.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\CD406B9D61C4.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\CD406B9D61CE.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\D0F6DE661D63.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\D0F6DE661D64.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\D0F6DE661D65.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\D28CDF85BDA3.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\D91DD06F9B3D.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\D91DD06F9B3E.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\DFC9150DB405.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\E04FFD273493.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\E04FFD273499.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\E3084BAFECCD.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\E89EFB01CEB9.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\EB2EFD28E53E.jpg\n",
      "E:\\SigTuple_Hack\\Test_Data\\F1BFEA74B33D.jpg\n"
     ]
    }
   ],
   "source": [
    "for img_file in ip_files:\n",
    "    input_file = 'E:\\\\SigTuple_Hack\\\\Test_Data\\\\' + img_file\n",
    "    mask_file = 'E:\\\\SigTuple_Hack\\\\Test_Data\\\\Mask\\\\' + img_file[:-4] + '-mask.jpg'\n",
    "    print input_file\n",
    "    \n",
    "    # Load the image & flatten 3d to 2d .\n",
    "    imgcolor = imread(input_file)/255.0\n",
    "    imgshape = imgcolor.shape\n",
    "    imgdf = pd.DataFrame(imgcolor.transpose(2,0,1).reshape(3,-1).transpose(1,0))\n",
    "    imgdf.columns = ['Red', 'Green', 'Blue']\n",
    "    \n",
    "    # Create mask, using LinearSVC model\n",
    "    mask = lin_clf.predict(imgdf)\n",
    "    \n",
    "    imgmask = mask.reshape(imgshape[0], imgshape[1])\n",
    "    \n",
    "    # Save file\n",
    "    scipy.misc.imsave(mask_file, imgmask)\n",
    "    \n",
    "    "
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}