{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f3c92fc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "from library.exercise import Exercise\n",
    "from library.dataset import DataSet, TrainTestData\n",
    "from library.GanExamples import StupidToyListGan\n",
    "from library.SimpleGan import SimpleGan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d10f9c73",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "import numpy as np\n",
    "import random\n",
    "from imblearn.datasets import fetch_datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "bfdaaf87",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'folding': [(array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.48, 0.45, 0.59, ..., 0.  , 0.58, 0.34],\n",
      "       [0.66, 0.55, 0.45, ..., 0.  , 0.46, 0.22],\n",
      "       [0.58, 0.47, 0.54, ..., 0.  , 0.51, 0.26],\n",
      "       ...,\n",
      "       [0.59, 0.67, 0.54, ..., 0.  , 0.48, 0.6 ],\n",
      "       [0.71, 0.5 , 0.5 , ..., 0.  , 0.46, 0.22],\n",
      "       [0.38, 0.4 , 0.66, ..., 0.  , 0.43, 0.11]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       [0.42, 0.44, 0.48, ..., 0.  , 0.48, 0.22],\n",
      "       [0.51, 0.4 , 0.56, ..., 0.  , 0.49, 0.22],\n",
      "       ...,\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.39, 0.58, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.63, 0.57, 0.5 , ..., 0.  , 0.51, 0.22],\n",
      "       [0.61, 0.48, 0.54, ..., 0.  , 0.5 , 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 1., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.81, 0.62, 0.43, ..., 0.  , 0.53, 0.22],\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.42, 0.44, 0.48, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.39, 0.54, ..., 0.  , 0.53, 0.27],\n",
      "       [0.42, 0.37, 0.59, ..., 0.  , 0.52, 0.29],\n",
      "       ...,\n",
      "       [0.48, 0.61, 0.57, ..., 0.  , 0.45, 0.22],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.81, 0.62, 0.43, ..., 0.  , 0.53, 0.22],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       [0.51, 0.4 , 0.56, ..., 0.  , 0.49, 0.22],\n",
      "       [0.5 , 0.54, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.46, 0.62, 0.46, ..., 0.  , 0.51, 0.27],\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.55, 0.5 , 0.66, ..., 0.  , 0.49, 0.22],\n",
      "       [0.4 , 0.39, 0.6 , ..., 0.  , 0.58, 0.3 ],\n",
      "       [0.61, 0.6 , 0.55, ..., 0.  , 0.5 , 0.25],\n",
      "       ...,\n",
      "       [0.6 , 0.61, 0.54, ..., 0.  , 0.46, 0.22],\n",
      "       [0.38, 0.32, 0.64, ..., 0.  , 0.44, 0.11],\n",
      "       [0.81, 0.62, 0.43, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       [0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       ...,\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.46, 0.44, 0.52, ..., 0.  , 0.5 , 0.22],\n",
      "       [0.47, 0.39, 0.5 , ..., 0.  , 0.49, 0.4 ],\n",
      "       ...,\n",
      "       [0.38, 0.47, 0.47, ..., 0.  , 0.44, 0.26],\n",
      "       [0.59, 0.67, 0.54, ..., 0.  , 0.48, 0.6 ],\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       [0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       ...,\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.42, 0.44, 0.48, ..., 0.  , 0.48, 0.22],\n",
      "       [0.55, 0.5 , 0.66, ..., 0.  , 0.49, 0.22],\n",
      "       ...,\n",
      "       [0.63, 0.57, 0.5 , ..., 0.  , 0.51, 0.22],\n",
      "       [0.48, 0.61, 0.57, ..., 0.  , 0.45, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       [0.4 , 0.39, 0.6 , ..., 0.  , 0.58, 0.3 ],\n",
      "       [0.58, 0.47, 0.54, ..., 0.  , 0.51, 0.26],\n",
      "       ...,\n",
      "       [0.38, 0.32, 0.64, ..., 0.  , 0.44, 0.11],\n",
      "       [0.38, 0.4 , 0.66, ..., 0.  , 0.43, 0.11],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 1.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.51, 0.4 , 0.56, ..., 0.  , 0.49, 0.22],\n",
      "       [0.5 , 0.54, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.51, 0.51, 0.52, ..., 0.  , 0.54, 0.22],\n",
      "       ...,\n",
      "       [0.58, 0.56, 0.38, ..., 0.  , 0.54, 0.57],\n",
      "       [0.6 , 0.61, 0.54, ..., 0.  , 0.46, 0.22],\n",
      "       [0.61, 0.48, 0.54, ..., 0.  , 0.5 , 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 1., 1., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       ...,\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       [0.48, 0.45, 0.59, ..., 0.  , 0.58, 0.34],\n",
      "       [0.61, 0.6 , 0.55, ..., 0.  , 0.5 , 0.25],\n",
      "       ...,\n",
      "       [0.25, 0.4 , 0.52, ..., 0.  , 0.56, 0.28],\n",
      "       [0.81, 0.62, 0.43, ..., 0.  , 0.53, 0.22],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.42, 0.44, 0.48, ..., 0.  , 0.48, 0.22],\n",
      "       ...,\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       [0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       [0.4 , 0.39, 0.6 , ..., 0.  , 0.58, 0.3 ],\n",
      "       ...,\n",
      "       [0.38, 0.47, 0.47, ..., 0.  , 0.44, 0.26],\n",
      "       [0.61, 0.48, 0.54, ..., 0.  , 0.5 , 0.22],\n",
      "       [0.38, 0.4 , 0.66, ..., 0.  , 0.43, 0.11]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       [0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       ...,\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.5 , 0.54, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.57, 0.47, 0.6 , ..., 0.  , 0.51, 0.22],\n",
      "       ...,\n",
      "       [0.51, 0.45, 0.62, ..., 0.  , 0.59, 0.22],\n",
      "       [0.48, 0.61, 0.57, ..., 0.  , 0.45, 0.22],\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       [0.58, 0.44, 0.57, ..., 0.  , 0.54, 0.22],\n",
      "       ...,\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.42, 0.44, 0.48, ..., 0.  , 0.48, 0.22],\n",
      "       [0.51, 0.4 , 0.56, ..., 0.  , 0.49, 0.22],\n",
      "       ...,\n",
      "       [0.71, 0.5 , 0.5 , ..., 0.  , 0.46, 0.22],\n",
      "       [0.38, 0.32, 0.64, ..., 0.  , 0.44, 0.11],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.4 , 0.42, 0.57, ..., 0.  , 0.53, 0.25],\n",
      "       [0.6 , 0.4 , 0.52, ..., 0.  , 0.53, 0.22],\n",
      "       [0.61, 0.6 , 0.55, ..., 0.  , 0.5 , 0.25],\n",
      "       ...,\n",
      "       [0.32, 0.29, 0.29, ..., 0.  , 0.52, 0.22],\n",
      "       [0.6 , 0.61, 0.54, ..., 0.  , 0.46, 0.22],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0.,\n",
      "       0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0.])), (array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
      "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
      "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
      "       ...,\n",
      "       [0.47, 0.43, 0.61, ..., 0.  , 0.48, 0.47],\n",
      "       [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
      "       [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]), array([0., 0., 0., ..., 0., 0., 0.]), array([[0.42, 0.37, 0.59, ..., 0.  , 0.52, 0.29],\n",
      "       [0.66, 0.55, 0.45, ..., 0.  , 0.46, 0.22],\n",
      "       [0.58, 0.47, 0.54, ..., 0.  , 0.51, 0.26],\n",
      "       ...,\n",
      "       [0.59, 0.67, 0.54, ..., 0.  , 0.48, 0.6 ],\n",
      "       [0.81, 0.62, 0.43, ..., 0.  , 0.53, 0.22],\n",
      "       [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39]]), array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.,\n",
      "       1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "       0., 0., 0., 0., 0., 0., 0.]))], 'db_size': 1484, 'db_n_attr': 10, 'imbalanced_ratio': 41.4}\n"
     ]
    }
   ],
   "source": [
    "datasetName = \"folding_yeast6\"\n",
    "pickle_in = open(f\"{datasetName}.pickle\",\"rb\")\n",
    "example_dict1 = pickle.load(pickle_in)\n",
    "\n",
    "print(example_dict1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "0f6b23e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['folding', 'db_size', 'db_n_attr', 'imbalanced_ratio'])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "example_dict1.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ce509dbc",
   "metadata": {},
   "outputs": [],
   "source": [
    "myData = example_dict1[\"folding\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "0fabea38",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
       "        [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
       "        [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
       "        ...,\n",
       "        [0.67, 0.57, 0.36, ..., 0.  , 0.56, 0.22],\n",
       "        [0.43, 0.4 , 0.6 , ..., 0.  , 0.53, 0.39],\n",
       "        [0.65, 0.54, 0.54, ..., 0.  , 0.53, 0.22]]),\n",
       " array([0., 0., 0., ..., 0., 0., 0.]),\n",
       " array([[0.48, 0.45, 0.59, ..., 0.  , 0.58, 0.34],\n",
       "        [0.66, 0.55, 0.45, ..., 0.  , 0.46, 0.22],\n",
       "        [0.58, 0.47, 0.54, ..., 0.  , 0.51, 0.26],\n",
       "        ...,\n",
       "        [0.59, 0.67, 0.54, ..., 0.  , 0.48, 0.6 ],\n",
       "        [0.71, 0.5 , 0.5 , ..., 0.  , 0.46, 0.22],\n",
       "        [0.38, 0.4 , 0.66, ..., 0.  , 0.43, 0.11]]),\n",
       " array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0.]))"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "myData[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4ff91174",
   "metadata": {},
   "outputs": [],
   "source": [
    "k = myData[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "9e31bb48",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_con = np.concatenate((k[0],k[2]),axis = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "727d84dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "target = np.concatenate((k[1],k[3]),axis = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "44270a01",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 0., 0., ..., 0., 0., 0.])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c95f51f4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.58, 0.61, 0.47, ..., 0.  , 0.48, 0.22],\n",
       "       [0.43, 0.67, 0.48, ..., 0.  , 0.53, 0.22],\n",
       "       [0.64, 0.62, 0.49, ..., 0.  , 0.53, 0.22],\n",
       "       ...,\n",
       "       [0.59, 0.67, 0.54, ..., 0.  , 0.48, 0.6 ],\n",
       "       [0.71, 0.5 , 0.5 , ..., 0.  , 0.46, 0.22],\n",
       "       [0.38, 0.4 , 0.66, ..., 0.  , 0.43, 0.11]])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_con"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "08c19220",
   "metadata": {},
   "outputs": [],
   "source": [
    "def loadDataset():\n",
    "    labels = target\n",
    "    features = data_con\n",
    "    label_1 = list(np.where(labels == 1)[0])\n",
    "    label_0 = list(np.where(labels == 0)[0])\n",
    "    features_1 = features[label_1]\n",
    "    features_0 = features[label_0]\n",
    "    \n",
    "    data0 = np.array(range(100)).reshape((25,4))\n",
    "    return DataSet(data0=features_0, data1=features_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5890f2ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = loadDataset()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "f27cef3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def getRandGen(initValue, incValue=257, multValue=101, modulus=65537):\n",
    "    value = initValue\n",
    "    while True:\n",
    "        value = ((multValue * value) + incValue) % modulus\n",
    "        yield value\n",
    "            \n",
    "randGen = getRandGen(2021)\n",
    "random.seed(2021)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "0939e1eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "def shuffler(data):\n",
    "    data = list(data)\n",
    "    size = len(data)\n",
    "    shuffled = []\n",
    "    while size > 0:\n",
    "        p = next(randGen) % size\n",
    "        size -= 1\n",
    "        shuffled.append(data[p])\n",
    "        data = data[0:p] + data[(p + 1):]\n",
    "    return np.array(shuffled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "32cf8258",
   "metadata": {},
   "outputs": [],
   "source": [
    "ganName = \"SimpleGAN\"\n",
    "gan = SimpleGan(numOfFeatures=k[2].shape[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "40220314",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> Shuffling data\n",
      "### Start exercise for synthetic point generator\n",
      "\n",
      "====== Step 1/5 =======\n",
      "-> Shuffling data\n",
      "-> Spliting data to slices\n",
      "\n",
      "------ Step 1/5: Slice 1/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 1\n",
      "LR fn, tp: 4, 3\n",
      "LR f1 score: 0.545\n",
      "LR cohens kappa score: 0.538\n",
      "LR average precision score: 0.635\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 287, 3\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.462\n",
      "GB cohens kappa score: 0.450\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 5, 2\n",
      "KNN f1 score: 0.444\n",
      "KNN cohens kappa score: 0.439\n",
      "\n",
      "\n",
      "------ Step 1/5: Slice 2/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 285, 5\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.286\n",
      "LR cohens kappa score: 0.268\n",
      "LR average precision score: 0.433\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 1\n",
      "GB fn, tp: 5, 2\n",
      "GB f1 score: 0.400\n",
      "GB cohens kappa score: 0.391\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 287, 3\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.462\n",
      "KNN cohens kappa score: 0.450\n",
      "\n",
      "\n",
      "------ Step 1/5: Slice 3/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 1\n",
      "LR fn, tp: 6, 1\n",
      "LR f1 score: 0.222\n",
      "LR cohens kappa score: 0.214\n",
      "LR average precision score: 0.455\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 1\n",
      "GB fn, tp: 5, 2\n",
      "GB f1 score: 0.400\n",
      "GB cohens kappa score: 0.391\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 288, 2\n",
      "KNN fn, tp: 6, 1\n",
      "KNN f1 score: 0.200\n",
      "KNN cohens kappa score: 0.189\n",
      "\n",
      "\n",
      "------ Step 1/5: Slice 4/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 1\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.400\n",
      "LR cohens kappa score: 0.391\n",
      "LR average precision score: 0.618\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 286, 4\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.429\n",
      "GB cohens kappa score: 0.415\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 5, 2\n",
      "KNN f1 score: 0.444\n",
      "KNN cohens kappa score: 0.439\n",
      "\n",
      "\n",
      "------ Step 1/5: Slice 5/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1132 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 287, 2\n",
      "LR fn, tp: 3, 4\n",
      "LR f1 score: 0.615\n",
      "LR cohens kappa score: 0.607\n",
      "LR average precision score: 0.662\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 288, 1\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.545\n",
      "GB cohens kappa score: 0.537\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 0\n",
      "KNN fn, tp: 1, 6\n",
      "KNN f1 score: 0.923\n",
      "KNN cohens kappa score: 0.921\n",
      "\n",
      "\n",
      "====== Step 2/5 =======\n",
      "-> Shuffling data\n",
      "-> Spliting data to slices\n",
      "\n",
      "------ Step 2/5: Slice 1/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 1\n",
      "LR fn, tp: 3, 4\n",
      "LR f1 score: 0.667\n",
      "LR cohens kappa score: 0.660\n",
      "LR average precision score: 0.677\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 287, 3\n",
      "GB fn, tp: 3, 4\n",
      "GB f1 score: 0.571\n",
      "GB cohens kappa score: 0.561\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 288, 2\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.500\n",
      "KNN cohens kappa score: 0.490\n",
      "\n",
      "\n",
      "------ Step 2/5: Slice 2/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 286, 4\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.308\n",
      "LR cohens kappa score: 0.292\n",
      "LR average precision score: 0.403\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 288, 2\n",
      "GB fn, tp: 5, 2\n",
      "GB f1 score: 0.364\n",
      "GB cohens kappa score: 0.353\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 5, 2\n",
      "KNN f1 score: 0.444\n",
      "KNN cohens kappa score: 0.439\n",
      "\n",
      "\n",
      "------ Step 2/5: Slice 3/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 1\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.400\n",
      "LR cohens kappa score: 0.391\n",
      "LR average precision score: 0.511\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 286, 4\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.429\n",
      "GB cohens kappa score: 0.415\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.600\n",
      "KNN cohens kappa score: 0.594\n",
      "\n",
      "\n",
      "------ Step 2/5: Slice 4/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 288, 2\n",
      "LR fn, tp: 4, 3\n",
      "LR f1 score: 0.500\n",
      "LR cohens kappa score: 0.490\n",
      "LR average precision score: 0.560\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 285, 5\n",
      "GB fn, tp: 5, 2\n",
      "GB f1 score: 0.286\n",
      "GB cohens kappa score: 0.268\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 288, 2\n",
      "KNN fn, tp: 5, 2\n",
      "KNN f1 score: 0.364\n",
      "KNN cohens kappa score: 0.353\n",
      "\n",
      "\n",
      "------ Step 2/5: Slice 5/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1132 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 288, 1\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.400\n",
      "LR cohens kappa score: 0.391\n",
      "LR average precision score: 0.507\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 0\n",
      "GB fn, tp: 5, 2\n",
      "GB f1 score: 0.444\n",
      "GB cohens kappa score: 0.439\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 0\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.600\n",
      "KNN cohens kappa score: 0.594\n",
      "\n",
      "\n",
      "====== Step 3/5 =======\n",
      "-> Shuffling data\n",
      "-> Spliting data to slices\n",
      "\n",
      "------ Step 3/5: Slice 1/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 1\n",
      "LR fn, tp: 4, 3\n",
      "LR f1 score: 0.545\n",
      "LR cohens kappa score: 0.538\n",
      "LR average precision score: 0.646\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 288, 2\n",
      "GB fn, tp: 3, 4\n",
      "GB f1 score: 0.615\n",
      "GB cohens kappa score: 0.607\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 1\n",
      "KNN fn, tp: 3, 4\n",
      "KNN f1 score: 0.667\n",
      "KNN cohens kappa score: 0.660\n",
      "\n",
      "\n",
      "------ Step 3/5: Slice 2/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 290, 0\n",
      "LR fn, tp: 3, 4\n",
      "LR f1 score: 0.727\n",
      "LR cohens kappa score: 0.723\n",
      "LR average precision score: 0.791\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 1\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.545\n",
      "GB cohens kappa score: 0.538\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.600\n",
      "KNN cohens kappa score: 0.594\n",
      "\n",
      "\n",
      "------ Step 3/5: Slice 3/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 288, 2\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.364\n",
      "LR cohens kappa score: 0.353\n",
      "LR average precision score: 0.451\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 1\n",
      "GB fn, tp: 6, 1\n",
      "GB f1 score: 0.222\n",
      "GB cohens kappa score: 0.214\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.600\n",
      "KNN cohens kappa score: 0.594\n",
      "\n",
      "\n",
      "------ Step 3/5: Slice 4/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 285, 5\n",
      "LR fn, tp: 3, 4\n",
      "LR f1 score: 0.500\n",
      "LR cohens kappa score: 0.486\n",
      "LR average precision score: 0.452\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 280, 10\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.300\n",
      "GB cohens kappa score: 0.278\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 285, 5\n",
      "KNN fn, tp: 3, 4\n",
      "KNN f1 score: 0.500\n",
      "KNN cohens kappa score: 0.486\n",
      "\n",
      "\n",
      "------ Step 3/5: Slice 5/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1132 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 0\n",
      "LR fn, tp: 7, 0\n",
      "LR f1 score: 0.000\n",
      "LR cohens kappa score: 0.000\n",
      "LR average precision score: 0.358\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 288, 1\n",
      "GB fn, tp: 7, 0\n",
      "GB f1 score: 0.000\n",
      "GB cohens kappa score: -0.006\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 0\n",
      "KNN fn, tp: 6, 1\n",
      "KNN f1 score: 0.250\n",
      "KNN cohens kappa score: 0.246\n",
      "\n",
      "\n",
      "====== Step 4/5 =======\n",
      "-> Shuffling data\n",
      "-> Spliting data to slices\n",
      "\n",
      "------ Step 4/5: Slice 1/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 290, 0\n",
      "LR fn, tp: 6, 1\n",
      "LR f1 score: 0.250\n",
      "LR cohens kappa score: 0.246\n",
      "LR average precision score: 0.726\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 1\n",
      "GB fn, tp: 2, 5\n",
      "GB f1 score: 0.769\n",
      "GB cohens kappa score: 0.764\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 288, 2\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.500\n",
      "KNN cohens kappa score: 0.490\n",
      "\n",
      "\n",
      "------ Step 4/5: Slice 2/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 286, 4\n",
      "LR fn, tp: 6, 1\n",
      "LR f1 score: 0.167\n",
      "LR cohens kappa score: 0.150\n",
      "LR average precision score: 0.263\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 286, 4\n",
      "GB fn, tp: 5, 2\n",
      "GB f1 score: 0.308\n",
      "GB cohens kappa score: 0.292\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 287, 3\n",
      "KNN fn, tp: 5, 2\n",
      "KNN f1 score: 0.333\n",
      "KNN cohens kappa score: 0.320\n",
      "\n",
      "\n",
      "------ Step 4/5: Slice 3/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 285, 5\n",
      "LR fn, tp: 3, 4\n",
      "LR f1 score: 0.500\n",
      "LR cohens kappa score: 0.486\n",
      "LR average precision score: 0.550\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 286, 4\n",
      "GB fn, tp: 3, 4\n",
      "GB f1 score: 0.533\n",
      "GB cohens kappa score: 0.521\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 1\n",
      "KNN fn, tp: 3, 4\n",
      "KNN f1 score: 0.667\n",
      "KNN cohens kappa score: 0.660\n",
      "\n",
      "\n",
      "------ Step 4/5: Slice 4/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 290, 0\n",
      "LR fn, tp: 4, 3\n",
      "LR f1 score: 0.600\n",
      "LR cohens kappa score: 0.594\n",
      "LR average precision score: 0.670\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 288, 2\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.500\n",
      "GB cohens kappa score: 0.490\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.600\n",
      "KNN cohens kappa score: 0.594\n",
      "\n",
      "\n",
      "------ Step 4/5: Slice 5/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1132 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 0\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.444\n",
      "LR cohens kappa score: 0.439\n",
      "LR average precision score: 0.587\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 287, 2\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.500\n",
      "GB cohens kappa score: 0.490\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 0\n",
      "KNN fn, tp: 5, 2\n",
      "KNN f1 score: 0.444\n",
      "KNN cohens kappa score: 0.439\n",
      "\n",
      "\n",
      "====== Step 5/5 =======\n",
      "-> Shuffling data\n",
      "-> Spliting data to slices\n",
      "\n",
      "------ Step 5/5: Slice 1/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 288, 2\n",
      "LR fn, tp: 3, 4\n",
      "LR f1 score: 0.615\n",
      "LR cohens kappa score: 0.607\n",
      "LR average precision score: 0.501\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 287, 3\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.462\n",
      "GB cohens kappa score: 0.450\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 287, 3\n",
      "KNN fn, tp: 3, 4\n",
      "KNN f1 score: 0.571\n",
      "KNN cohens kappa score: 0.561\n",
      "\n",
      "\n",
      "------ Step 5/5: Slice 2/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 289, 1\n",
      "LR fn, tp: 6, 1\n",
      "LR f1 score: 0.222\n",
      "LR cohens kappa score: 0.214\n",
      "LR average precision score: 0.236\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 288, 2\n",
      "GB fn, tp: 5, 2\n",
      "GB f1 score: 0.364\n",
      "GB cohens kappa score: 0.353\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 6, 1\n",
      "KNN f1 score: 0.250\n",
      "KNN cohens kappa score: 0.246\n",
      "\n",
      "\n",
      "------ Step 5/5: Slice 3/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 290, 0\n",
      "LR fn, tp: 4, 3\n",
      "LR f1 score: 0.600\n",
      "LR cohens kappa score: 0.594\n",
      "LR average precision score: 0.744\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 1\n",
      "GB fn, tp: 1, 6\n",
      "GB f1 score: 0.857\n",
      "GB cohens kappa score: 0.854\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 1\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.545\n",
      "KNN cohens kappa score: 0.538\n",
      "\n",
      "\n",
      "------ Step 5/5: Slice 4/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1131 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 288, 2\n",
      "LR fn, tp: 5, 2\n",
      "LR f1 score: 0.364\n",
      "LR cohens kappa score: 0.353\n",
      "LR average precision score: 0.542\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 289, 1\n",
      "GB fn, tp: 4, 3\n",
      "GB f1 score: 0.545\n",
      "GB cohens kappa score: 0.538\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 290, 0\n",
      "KNN fn, tp: 5, 2\n",
      "KNN f1 score: 0.444\n",
      "KNN cohens kappa score: 0.439\n",
      "\n",
      "\n",
      "------ Step 5/5: Slice 5/5 -------\n",
      "-> Reset the GAN\n",
      "-> Train generator for synthetic samples\n",
      "Epoch 1/3\n",
      "Epoch 2/3\n",
      "Epoch 3/3\n",
      "-> create 1132 synthetic samples\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> test with 'LR'\n",
      "LR tn, fp: 285, 4\n",
      "LR fn, tp: 4, 3\n",
      "LR f1 score: 0.429\n",
      "LR cohens kappa score: 0.415\n",
      "LR average precision score: 0.443\n",
      "\n",
      "-> test with 'GB'\n",
      "GB tn, fp: 287, 2\n",
      "GB fn, tp: 6, 1\n",
      "GB f1 score: 0.200\n",
      "GB cohens kappa score: 0.188\n",
      "\n",
      "-> test with 'KNN'\n",
      "KNN tn, fp: 289, 0\n",
      "KNN fn, tp: 4, 3\n",
      "KNN f1 score: 0.600\n",
      "KNN cohens kappa score: 0.594\n",
      "\n",
      "### Exercise is done.\n",
      "\n",
      "-----[ LR ]-----\n",
      "maximum:\n",
      "LR tn, fp: 290, 5\n",
      "LR fn, tp: 7, 4\n",
      "LR f1 score: 0.727\n",
      "LR cohens kappa score: 0.723\n",
      "LR average precision score: 0.791\n",
      "\n",
      "\n",
      "average:\n",
      "LR tn, fp: 288.0, 1.8\n",
      "LR fn, tp: 4.52, 2.48\n",
      "LR f1 score: 0.427\n",
      "LR cohens kappa score: 0.418\n",
      "LR average precision score: 0.537\n",
      "\n",
      "\n",
      "minimum:\n",
      "LR tn, fp: 285, 0\n",
      "LR fn, tp: 3, 0\n",
      "LR f1 score: 0.000\n",
      "LR cohens kappa score: 0.000\n",
      "LR average precision score: 0.236\n",
      "\n",
      "\n",
      "-----[ GB ]-----\n",
      "maximum:\n",
      "GB tn, fp: 289, 10\n",
      "GB fn, tp: 7, 6\n",
      "GB f1 score: 0.857\n",
      "GB cohens kappa score: 0.854\n",
      "\n",
      "\n",
      "average:\n",
      "GB tn, fp: 287.36, 2.44\n",
      "GB fn, tp: 4.24, 2.76\n",
      "GB f1 score: 0.442\n",
      "GB cohens kappa score: 0.432\n",
      "\n",
      "\n",
      "minimum:\n",
      "GB tn, fp: 280, 0\n",
      "GB fn, tp: 1, 0\n",
      "GB f1 score: 0.000\n",
      "GB cohens kappa score: -0.006\n",
      "\n",
      "\n",
      "-----[ KNN ]-----\n",
      "maximum:\n",
      "KNN tn, fp: 290, 5\n",
      "KNN fn, tp: 6, 6\n",
      "KNN f1 score: 0.923\n",
      "KNN cohens kappa score: 0.921\n",
      "\n",
      "\n",
      "average:\n",
      "KNN tn, fp: 288.8, 1.0\n",
      "KNN fn, tp: 4.24, 2.76\n",
      "KNN f1 score: 0.502\n",
      "KNN cohens kappa score: 0.495\n",
      "\n",
      "\n",
      "minimum:\n",
      "KNN tn, fp: 285, 0\n",
      "KNN fn, tp: 1, 1\n",
      "KNN f1 score: 0.200\n",
      "KNN cohens kappa score: 0.189\n",
      "\n"
     ]
    }
   ],
   "source": [
    "exercise = Exercise(shuffleFunction=shuffler, numOfShuffles=5, numOfSlices=5)\n",
    "exercise.run(gan, data)\n",
    "exercise.saveResultsTo(f\"{datasetName}-{ganName}.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3b6bac36",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}