generate_test_dataset.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from scipy.misc import imresize\n",
    "from imageio import imread\n",
    "import shutil\n",
    "import pandas as pd\n",
    "import pickle\n",
    "from tqdm import tqdm_notebook, tqdm\n",
    "from sklearn.model_selection import train_test_split\n",
    "import seaborn as sns\n",
    "from pylab import rcParams\n",
    "# rcParams['figure.figsize']=10,10\n",
    "import sys\n",
    "from sklearn.metrics import log_loss, accuracy_score, confusion_matrix\n",
    "import cv2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('data/x_test.pickle', 'rb') as f:\n",
    "    x_test=np.array(pickle.load(f))\n",
    "with open('data/y_test.pickle', 'rb') as f:\n",
    "    y_test=np.array(pickle.load(f))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['6e81892db1569782724a0234e610868a.png',\n",
       "       '0bcfc5a4d920dcced672ad82fa136875.png',\n",
       "       'c26d812059636151e2d7264fa0a9fdbc.png',\n",
       "       '0c6bdc9e7cc22fdd4266e39e58fd87be.png',\n",
       "       '5aced0ab3e06a9428cf54767cd5b8df6.png',\n",
       "       '3fc5e991d9d67ade69e95d8422b8b694.png',\n",
       "       '4a46f53037b474194327f458ac83d5ad.png',\n",
       "       '1bd0022f266d1a671a126e565318e12a.png',\n",
       "       '5f2c3de190f1e231fbb68d64eae7badc.png',\n",
       "       'b4d425eedb47b3d886ca70d56db4120d.png',\n",
       "       '0bd52690187d601455f5b83670fe019f.png',\n",
       "       '0e672b307e510b033d601b4901ca9a4c.png',\n",
       "       '0ce3e0b2cb776b958bb763f0ba8c46f9.png',\n",
       "       '3ee46c9946fd48e2cdf159e0dd42dd38.png',\n",
       "       'd8edc14c38478d80b27a0bc6d41dc4ae.png',\n",
       "       'b318ed71fdddcabf17e15c6ca3af13af.png',\n",
       "       '4a28995506ac48854eb426a5e06c12d0.png',\n",
       "       '6d33f983240a4a474d51cd32d05e918d.png',\n",
       "       '2fbab30ef74dd450672d2a9d3142b459.png',\n",
       "       '4fbf3c5ad32f5e222ed1953316ce0d51.png',\n",
       "       'a3978283d13463af3d6d98c6ae6d6be7.png',\n",
       "       'afe58b89ed86f93d38c8dd156a8a8491.png',\n",
       "       '4bc96e868d4eb0477881c14c65c112a4.png',\n",
       "       '4a694369bbb2e7271c7783d79f77e862.png',\n",
       "       '689303d2322edb19c223bc2fbe409e08.png',\n",
       "       '0c420e5d4ad357c01f6f16d9d4c9a5c5.png',\n",
       "       '6cd39a1288e2650d1c8ebafde7d9d103.png',\n",
       "       '1fc25a4373e329f72e8770b9446ec0a0.png',\n",
       "       'b957bf062b8006317f24bdba5dd0abfe.png',\n",
       "       'b35d925d02a6792e66e475860372aaae.png',\n",
       "       '1f51bb9d9f79557fcedfef1d3bc931bc.png',\n",
       "       '0c43945b3bd42eb2f0e68a7f7ed43db0.png',\n",
       "       'e3f297c613666f68896094cb58838ae6.png',\n",
       "       '0bad6f172bb05aa05f2ad785c0f7cb9f.png',\n",
       "       '7a303fa7d3ed5b2f4bbe3f0772fd2247.png',\n",
       "       '002e23c036a05de656242f5ce6ab2e15.png',\n",
       "       '4c15ec243178e6fea3390f17b1c0070c.png',\n",
       "       '5ef21f399c27230a5e4dfd9153d0fbe8.png',\n",
       "       '5ce799ff14670b83ad1a9e7ebb99ce6e.png',\n",
       "       '1d91a3f7e1d9799eeadc4fb572f5c344.png',\n",
       "       'b815a74cfdca7f5da57fc1b0687f89cd.png',\n",
       "       'bac7fb0f429c30179ca36411f11713ed.png',\n",
       "       'c6d6089b4ea367333ca025d8cd1c8d33.png',\n",
       "       '0eb173b4c0c9b6e8db3de6bcfafbc421.png',\n",
       "       '55774d613ff0e35e640172a35fdd6c96.png',\n",
       "       '6a5a499626b3e261fa2cb6cf15dcf59d.png',\n",
       "       '6fb1f8729a7064e55ba916655c2b347d.png',\n",
       "       '6d31084bb48b6fc5fbd6ff316459cd5b.png',\n",
       "       '7a61a6a3eda2192b5f03a5b54cb54654.png',\n",
       "       '6d969d0fe88819a9d0efd692454c5f7f.png',\n",
       "       'ccedb1eb235f5dedddc443b93e9aaabd.png',\n",
       "       'cc1234027ede8d9006ce01f0cd12b710.png',\n",
       "       'e81d4326a9aba0df4346edaf4e64c0d0.png',\n",
       "       '6b31c9e353b7135cf983f7f7cfe354ca.png',\n",
       "       'ce5cd74980494117a495a305ed5fe95b.png',\n",
       "       '729184490532feff1bea99534924c91d.png',\n",
       "       'c658e082c4a76075a2f91c731b1f79a0.png',\n",
       "       'b9149cc8a64a111bbc1b30cb1bdc37d4.png',\n",
       "       '2cfa2dc4c58aa3766f022100135a313a.png',\n",
       "       '6fcdaa82f912a1b56b3f24501a30aecc.png',\n",
       "       '1eb6e347587776b69c5a86cd5985e5c7.png',\n",
       "       'c78759a7396dd709da30a25683b6219c.png',\n",
       "       '6a8a86869eae63af596fd14e50f2b3c6.png',\n",
       "       'ae5cbfd5a2566eb2ba2e27979c53ef9c.png',\n",
       "       '6bf3f0d65bcb5bbc6f569147662966a4.png',\n",
       "       '5d150087e7982172941a6f0fbc12e3f1.png',\n",
       "       'cfb77b3822d05b280d51e8259f0c2ff4.png',\n",
       "       '14548214e97adf237ee3bac90809baf6.png',\n",
       "       '1ec8a66fae209ac812a39e11b033574b.png',\n",
       "       '0001d52e2fd94f30c2bca0449763a752.png',\n",
       "       '3a38f1fd519ebe4949980a4f4ab340a5.png',\n",
       "       '2d207c1b2bf4c522c5a19150ceab2829.png',\n",
       "       '0f215e9b7f0fbe5c6008d2f094d76df5.png',\n",
       "       '4dc002078cbb9dc4ee09935f6c0ac8c4.png',\n",
       "       '3e4860b730c20994297e663d93e2fb0f.png',\n",
       "       '5fb090a35bfa08776c41e854d6d274ca.png',\n",
       "       'af8332f8c5b8cedeeb33963bb532e2f8.png',\n",
       "       'b12fb5198adce69091a77caac294016c.png',\n",
       "       '1b9b2acb177b1c54a37e297aa89dac7a.png',\n",
       "       '4a7ab4c11dc562171cc0424372223830.png',\n",
       "       'b5413c246d39c53af7e53b1d56f64946.png',\n",
       "       '6a68286e27dacf32aeaf277854f493a6.png',\n",
       "       '1b3d595d91ba77637ac6ea6339af9af6.png',\n",
       "       '6e06c8209db6044dcd43d507896ceda5.png',\n",
       "       '03cf7a8ef349c3b618f95a85093568db.png',\n",
       "       '1b095a33602a6a21cee13c6d8c1415bc.png',\n",
       "       '0f745cf2b785e7a64620f09cf2590092.png',\n",
       "       'c0a07e462960bd1eb37b5b0c1753c0fa.png',\n",
       "       '3d59e2121cd4a713883a1b06cb7b11ab.png',\n",
       "       'b1a7bc9873d1ac1a9bf66426002b3722.png',\n",
       "       '6fa364a78c495fcf632f9539017ab50d.png',\n",
       "       '6bfdb3ff45f7307d29ea2ad92c4fbea9.png',\n",
       "       '2f88b06c9008b647bf4462240ad719bc.png',\n",
       "       '1c853fc69f2b4c662af17fd02be45828.png',\n",
       "       '1b2d8da303239dd9e189d58840825be5.png',\n",
       "       '0f1fe6db70e9da10f9653516e37e780a.png',\n",
       "       'e8540972740aa21638731940c3bebe42.png',\n",
       "       '1f20849903aa5109dc0eec33a2871840.png',\n",
       "       '5e66ff6f6051267582e73ff4190c0a79.png',\n",
       "       '06eefed61325a0c81ea9fed64f07d289.png',\n",
       "       '2377136d0e2ec5c8aa15293c78ffa3d7.png',\n",
       "       'd257675d358d70474a86b4e11dba33a9.png',\n",
       "       '3cfab460cd775b2fc3261f9f9d387c9e.png',\n",
       "       'd958cd2313d0c5eb6d55c9d96c25c8b8.png',\n",
       "       '6a486ced414771ef1f43b76f25ecb472.png',\n",
       "       'c2ccae30ccc8189fe03319b604cdde8b.png',\n",
       "       'b0060704d02f1229b75cbd550c7267b4.png',\n",
       "       '1d8f78e9c5b324fe80fad00cde8dd8d9.png',\n",
       "       '4f208c2befbfc8431aeb7f9e93d2bf64.png',\n",
       "       'ae6ec565eab53a8bdc5749cff5914726.png',\n",
       "       '5c5cca73226200f62643ccfa53282d90.png',\n",
       "       '2ba762dd3dad9a4704c46c8cf9a0ef13.png',\n",
       "       'b78a5e18d75253830e1c0cf7f73be279.png',\n",
       "       '2170480823e668068b78bce75afe07a1.png',\n",
       "       '6a486eca64c70c43eb1c9985458ffa97.png',\n",
       "       '3aad6b1da20e2215bbc44db86928791e.png',\n",
       "       '6afa3c94f2e59d887d5a0df6359dee11.png',\n",
       "       'd22f7eb37563c080925a4cacbed16c7f.png',\n",
       "       'ca76eefff717b507a0eba5353324d9e5.png',\n",
       "       '5c053d4e9065093c17dc5ca9154b6577.png',\n",
       "       '4f15fcfaea4ca11088275f4377704d6a.png',\n",
       "       '00e3659be9c1fec47c739f79252840e3.png',\n",
       "       '4e8df12016701a5eb8fabdb728877c26.png',\n",
       "       '4a1aa0cab7646c49d647bd8e4d975190.png',\n",
       "       '00fbcae4d05aeab62ab917e587b911f9.png',\n",
       "       '022354380a1a17ff49226c9eae21a11f.png',\n",
       "       '5c3fef9b45dc171c7c082e12b7f6eb5c.png',\n",
       "       'ac1a667900157fae86ec5fdabe5eb4b0.png',\n",
       "       '2cbfe7b264f0671635ff4557cb346725.png',\n",
       "       '6be828b2819012072acadc613f81e927.png',\n",
       "       '6aac18ba35542f2a9fd6d91d7e202549.png',\n",
       "       '7adb7a9c863b8bdd842d14e1c484e77f.png',\n",
       "       '0d20e86dad11f0bd9ccb6ffa014277f2.png',\n",
       "       'a470313562508266ecf17a4a5410fc6f.png',\n",
       "       'add77a1c9b7aa95a65542a01a209b7d2.png',\n",
       "       'e44cb44a7cf02f5c4f9f4c2f0f62f3ce.png',\n",
       "       '2d68cfb6b98a0aba7781e10d8264a3be.png',\n",
       "       '1c7d4d157e04b416d2d2924abfa08dc8.png',\n",
       "       '4cfce60b3c6e30635dbcbae55d9e5f01.png',\n",
       "       '4ec11a3e90f698245558752bba778c6e.png',\n",
       "       '152681a0017a5fded699c43cd6df97d1.png',\n",
       "       '0f68b5903eb36be82019631a333c9141.png',\n",
       "       'b47c970b9a25c103951ab48c55727ecc.png',\n",
       "       'ca8d2d96f0802516c94fa60dbae9e0eb.png',\n",
       "       'd38f15e7ee1656b427fc23cea9944691.png',\n",
       "       '0c0b302cb4a6f593cbb34b803b6ad71f.png',\n",
       "       '0e429185c58e9dfcaa3cfee8bbf1e830.png',\n",
       "       '6b3eb059fd9044414327cfe2db54b1a0.png',\n",
       "       '2a9b0535dc7182014752ec66fa65a2d0.png',\n",
       "       '04f3c84abe8419ead6de6515a2dc05de.png',\n",
       "       '6fbecfdaecf314fc8417362440c1867c.png',\n",
       "       'cd9b2df963652cea69a16b7e368cd7f1.png',\n",
       "       'c834e3170c81cebd468accb10ef07c1f.png',\n",
       "       '0a44837310aaf998acb05d0a651891e3.png',\n",
       "       '3ddf7736a773413d96643ce52c9a3c52.png',\n",
       "       '4ec45a6cecd8a24a5ec3310ce19b2d73.png',\n",
       "       '4bc7e3fe32e048305059534f32b924e2.png',\n",
       "       '6b81b4828b7bd9fd450258cc233ec70f.png',\n",
       "       '05d0476ae1a480e1a79e7cfae6ab3a5d.png',\n",
       "       'c4249fa108dfc8cc6617d0c03adaa9e4.png',\n",
       "       '1a17a41f584faa97befe50fdf72ba13d.png',\n",
       "       '3c4141c3fb510f0077807d94aa1a7e7b.png',\n",
       "       '06c190007f980b9d58a25a6db433327a.png',\n",
       "       'a464044a6cdff5cff44d099a99138ae2.png',\n",
       "       '5fcbff0dea2096296955f621ea2ed24b.png',\n",
       "       '1c7fbbb2b22971f0e6afdc9a0bc6bd46.png',\n",
       "       '3cb69784be21bce972537f1ebf0bb231.png',\n",
       "       '1ebbb5a00989068e50000aaa5d5dd5fc.png',\n",
       "       '3c61021caf93f68df55053972dce9394.png',\n",
       "       'c9df6c726d44834138a89684b2454dbb.png',\n",
       "       'd87fb93838434f8a1d668c0ed1bd8824.png',\n",
       "       '2eb515a0b14569b9c98f3cb7e0a62686.png',\n",
       "       '03eaa5163b84efa2b14c5d1431107200.png',\n",
       "       '6a26bca72eec83c3eb6290b471f76d72.png',\n",
       "       '3caa8e3dbe2f3572dfb66cfea95a729e.png',\n",
       "       '3b915720d58708f83561ade96107afa6.png',\n",
       "       'b42396107373446a2ab013cd61460b04.png',\n",
       "       '3c8c56e88b6aebb17800b9b27ea69969.png',\n",
       "       '3deb3cb2c8f1d3179dcb8ba43f9df7db.png',\n",
       "       '5a33babc5fab4dfb80959bfa892f18b0.png',\n",
       "       '6c4886c8ae23964b735ceed09e39bb30.png',\n",
       "       '5f08f620b624894e6c48f938343861a1.png',\n",
       "       '05d9c3c9bb54b8beca02ac89553bd84a.png',\n",
       "       '1ad910c71cc71a500e14b68dd16ac00b.png',\n",
       "       'ac821291d07a9c317f2dbf5e424cb8cb.png',\n",
       "       '5c93378bb928ce87fe7613059848242d.png',\n",
       "       '3ce72f69f1d70fa6f4251bf620c971af.png',\n",
       "       '3a7f78afa41fc79c60edffe1e1853f06.png',\n",
       "       '2c9efd7686482f3fd1687c8bed7e1b79.png',\n",
       "       '4b27ce4bea6a576d928932f005e954cb.png',\n",
       "       'd3178109fc141fbce7c7d2c97a9e03d8.png',\n",
       "       '6e34a1f2e448d880b2d98fe7ecf2c9ec.png',\n",
       "       '0908dafde12041540b70d688315df6e9.png',\n",
       "       '6bd4cd258cf603a3563805dc08a61717.png',\n",
       "       '4c7cffa0740d46811611e1a1fb4a33dd.png',\n",
       "       '1f43b94aa7b64711ad40157a3b257a70.png',\n",
       "       '6e91f788bd7f03474fb5f7174ee3b4fe.png',\n",
       "       '4f70340d893fc1e430ea13c452516689.png',\n",
       "       '4a3c87626d683848038323bf2c172401.png',\n",
       "       '3bcf7a0569ed8c3e8ad875f95bfb9375.png',\n",
       "       '4afe9d1ba7e6165ec16f65480716f637.png',\n",
       "       '2abc3819330e8b6613596f603122f3f7.png',\n",
       "       '1d8ce4e2c38d5802f756fe1a295bafa7.png',\n",
       "       '1a9a240d5c549ea5f6a464f076d5f783.png',\n",
       "       '3c4c56e17f62c51c30f128f4d754f001.png',\n",
       "       'ba88fff70edc152e7fd26df38f710f49.png',\n",
       "       '3a13e041e01f6567262b899b5f2c5b9c.png',\n",
       "       '1fcc8711337c90f6c3dcfd1a07ed8a13.png',\n",
       "       '0fbb5ba2d58fdf6c4f129a876c7c0004.png',\n",
       "       '6bf06a5c16a66398eb1fe02d74abcb30.png',\n",
       "       '6743af3f663bc4244b0f80b93541f542.png',\n",
       "       '1ab03120be532e415d61102e9ce30edf.png',\n",
       "       '2d434fc0c848798124376812abd327c8.png',\n",
       "       'cdce0701fa96db632af5facf074fe654.png',\n",
       "       '04b9dd7d5e3d24e0bca41ff9173626f1.png',\n",
       "       '0bcebfad9022dc4e9b606d3bbe84483d.png',\n",
       "       '6a7ff6587ff54c98c4ae74e6919ba872.png',\n",
       "       '6d11eb7dfe32df914c7503fe6a41ce84.png',\n",
       "       '7a91051d33e92b96c9dd507eec4b8ca8.png',\n",
       "       '5a11ffa108c7b9400c1fc8b53242fe75.png',\n",
       "       '6b48a8f1eae42db610c632b7169b78a9.png',\n",
       "       '62181550c772f3277ca96e0516b45d9b.png',\n",
       "       'ac7136dac7ab3dfa07d4a6c5038f450d.png',\n",
       "       '4abc145709ef0a49fa40ab296d378977.png',\n",
       "       '3c0045d491e0b59c5727c5d59feb1064.png',\n",
       "       '6cf810778c1e0fd4d95cc5009d44f165.png',\n",
       "       '0bf69d1b9460a9ad974b9a7c57f31632.png',\n",
       "       'b0a2e857fb69d84cc86e16e53c2feda3.png',\n",
       "       '919881fcd4d0a4117719449a60788f18.png',\n",
       "       'cd23dec47118e009aa9ae014362be38f.png',\n",
       "       'bfd1934c21fe48a0e3ad8e9695473cb8.png',\n",
       "       '0f510d92d43b1b2f9e9ed73957cf87e8.png',\n",
       "       '5a3c4533dd0b0ae4b4dd7149fa2bf487.png',\n",
       "       '871d73986285b6092cc16cfe7ab4b7ed.png',\n",
       "       '4fedf7612fb6ea67a4a9b963396e3b2a.png',\n",
       "       '0e2f6aab6ade6e5b130064d6c2d79383.png',\n",
       "       '02efa34b5034a41fb487ef38f07c321c.png',\n",
       "       'a3902e315de94e425dfd3bd109dc88de.png',\n",
       "       'af8c8581458892ce989aaa67e65904c5.png',\n",
       "       '6b81062ee28cdb5ab08d5e86d4d20049.png',\n",
       "       '5a2c17fb2299adfb232fb5c99b48afbc.png',\n",
       "       '2c0402a97ce575430dd6ffaf023ee585.png',\n",
       "       '8569531f0cfe6fed6f0911100c8c8d56.png',\n",
       "       '5a5c1f2e74a9eb081113929fbac9e77f.png',\n",
       "       '4c62ff1dbfb7828b2432194ff0ce0b5a.png',\n",
       "       '07af4ffd4a3e7a37903b9a22affccfed.png',\n",
       "       '6870d1aa15aa6be76dac6931583bf998.png',\n",
       "       '2680cb774242658dd33dd7dccccf6308.png',\n",
       "       '5d7faad1f366ab67568d185a0dce8173.png',\n",
       "       '4d493266de1cefe31575bfa475839089.png',\n",
       "       '2a4df7ec33239ffbaf4f5b9a0b190ae3.png',\n",
       "       '6e5d61030136ad5ca49d667fa570e4fc.png',\n",
       "       '6dcfaabd782bd7dd76eb4e4acb634789.png',\n",
       "       '3cc9528009f979daf3681b63a47a85af.png',\n",
       "       'cb73316904b4dfa07e2f0066fdc91242.png',\n",
       "       'd12f8acae725c66b00c31c3c96978e18.png',\n",
       "       '5c59a6e897670b94b1d68faa23a8946a.png',\n",
       "       '4c10db55f29f713289bc4c52cc4e71f0.png',\n",
       "       '4ccb5bc9d04a0410769470637b9a8e6a.png',\n",
       "       '3a4890683ea8f09afd9a80f48686cef5.png',\n",
       "       'bc6699010f8031e38623087db86466b1.png',\n",
       "       '6b2a57077c246f5a96a9009dc9f86ab6.png',\n",
       "       'f95d94a3a6384059a64725f89677e885.png',\n",
       "       'd9b9f5db7d29a3855cceef574145b595.png',\n",
       "       '0e12b739725814023bf25b7c91780852.png',\n",
       "       'b2474c062971986f588459818c212a8c.png',\n",
       "       '2a9f3a3fec8d52af7cc2ceed1ca12c13.png',\n",
       "       '4db6603b65d03796d5ecde2a6a421f03.png',\n",
       "       '6cdc55e67e54628c981eba6be967f7e1.png',\n",
       "       '7afcfe686667df48e62e685663460846.png',\n",
       "       '1bbe5358b6911fd52ac6383a652292ee.png',\n",
       "       'b15e0636218f1748f362cb65648a2408.png',\n",
       "       'ce4df5feb17e4951e1151a4ddf38b5e5.png',\n",
       "       'bdaed1accdd28cf71be52229b654c0fd.png',\n",
       "       '3cfc9cc91c80cae3140e5c833ba03b6a.png',\n",
       "       '8403960a267cea6cec8473736454c4bd.png',\n",
       "       '3e2675aef79e152afb49b648c2892d8a.png',\n",
       "       '5b166a348430539eb9c9be6ccce20a0e.png',\n",
       "       '7a58f23c07cf45256f308b5723af9f7c.png',\n",
       "       '0c6ac3ce5ed62c27b99394e652c52ae0.png',\n",
       "       '0a8335445b959de1c8fac4360906d3b4.png',\n",
       "       'cebf26e8911e2a38c273b1c015e61fcd.png',\n",
       "       'cccbca215f38e9a03c46f73265bba889.png',\n",
       "       '6b5559315697d36ff30a772b77b1ef9c.png',\n",
       "       '2d7530637fce7dc799e344434cf2c1c1.png',\n",
       "       'd0c156d5cf865a9e43e20c1798093bf0.png',\n",
       "       '5da397367179d6cceb846dfbce4f99a5.png',\n",
       "       'a0ed4e883bd44522343ad589aadce538.png',\n",
       "       '0ce49d459921e224bba36d3b41c3c2da.png',\n",
       "       '3f730117c58c8e112a7b33f1331e9ef0.png',\n",
       "       'bc04da26ab41ce92565dd3c686dae6c8.png',\n",
       "       '5a0a481b717fd0e0f199589da2eb3883.png',\n",
       "       '02f58a2257e7aa18d2bb61deae00e6b5.png',\n",
       "       '2d2109bc991437a2e9b0edb0efdd1434.png',\n",
       "       '6ad2ca11ceb5ff971d4ae66220d71040.png'], dtype='<U36')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 1,\n",
       " 0,\n",
       " 0,\n",
       " 0,\n",
       " 0]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test_pristine=x_test[y_test==0]\n",
    "x_test_fake=x_test[y_test==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "90"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(x_test_fake)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test_fake = np.load('k64 binary 25percent stride8/sample_fakes_np_test.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51109"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(x_test_fake)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('data/x_test.pickle', 'rb') as f:\n",
    "    x_test=np.array(pickle.load(f))\n",
    "\n",
    "with open('data/y_test.pickle', 'rb') as f:\n",
    "    y_test=np.array(pickle.load(f))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test_pristine_names=x_test[y_test==0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sample_random(img, num_samples, stride=8):\n",
    "    kernel_size=64\n",
    "    \n",
    "    x_start=0\n",
    "    y_start=0\n",
    "    samples=[]\n",
    "    \n",
    "    for y_start in range(0, img.shape[0] - kernel_size + 1, stride):\n",
    "        for x_start in range(0, img.shape[1] - kernel_size + 1, stride):\n",
    "\n",
    "            #c_255 = count_255(mask[y_start:y_start + kernel_size, x_start:x_start + kernel_size])\n",
    "\n",
    "            #if (c_255 > threshold) and (kernel_size * kernel_size - c_255 > threshold):\n",
    "            samples.append(img[y_start:y_start + kernel_size, x_start:x_start + kernel_size, :3])\n",
    "\n",
    "    \n",
    "    indices=np.random.randint(0, len(samples), min(len(samples), num_samples))\n",
    "    \n",
    "    sampled=[]\n",
    "    for i in indices:\n",
    "        sampled.append(samples[i])\n",
    "    \n",
    "    return sampled"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "205"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(x_test_pristine_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "249.31219512195122"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(x_test_fake)/len(x_test_pristine_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "249 samples per pristine image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51045"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "249*205"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "fake_path = 'dataset-dist/phase-01/training/fake/'\n",
    "pristine_path = 'dataset-dist/phase-01/training/pristine/'\n",
    "mask_path = fake_path + 'masks/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test_pristine = []\n",
    "for img_name in x_test_pristine_names:\n",
    "    if os.path.isfile(pristine_path + img_name):\n",
    "        pristine_img = imread(pristine_path + img_name)\n",
    "\n",
    "\n",
    "        x_test_pristine.append(pristine_img[:, :, :3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "51045\n"
     ]
    }
   ],
   "source": [
    "samples_pristine_binary=np.ndarray(shape=(51045, 64, 64, 3), dtype=np.dtype('uint8'))\n",
    "\n",
    "i=0\n",
    "for pristine_img in x_test_pristine:\n",
    "    \n",
    "    samples=sample_random(pristine_img, 249)\n",
    "    for j, sample in enumerate(samples):\n",
    "        samples_pristine_binary[i, :, :, :]=sample\n",
    "        i+=1\n",
    "print(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('samples_pristine_test.npy', samples_pristine_binary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x208a2cdbfd0>"
      ]
     },
     "execution_count": 37,
     "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.imshow(samples_pristine_binary[13])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(51109, 64, 64, 3)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test_fake.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test=np.concatenate((x_test_fake, samples_pristine_binary), axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(102154, 64, 64, 3)"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('x_test.npy', x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test=np.array([1]*len(x_test_fake)+[0]*len(samples_pristine_binary))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('y_test.npy', y_test)"
   ]
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}