test_vision.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision import datasets, transforms\n",
    "import pickle\n",
    "from PIL import Image\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import hmax\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\arsen\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\numpy\\core\\fromnumeric.py:43: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  result = getattr(asarray(obj), method)(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "#test data download\n",
    "eye_movements_df = pd.read_excel('./data/eye_movements/sample_data1(dynamics).xlsx')\n",
    "\n",
    "x = np.array(eye_movements_df['Point of Regard Right X [px]'])\n",
    "y = np.array(eye_movements_df['Point of Regard Right Y [px]'])\n",
    "\n",
    "x = np.split(x, np.where(x == '-')[0])\n",
    "y = np.split(x, np.where(y == '-')[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = hmax.HMAX('./universal_patch_set.mat')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# A folder with example images\n",
    "example_images = datasets.ImageFolder(\n",
    "    './example_images/',\n",
    "    transform=transforms.Compose([\n",
    "        transforms.Grayscale(),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Lambda(lambda x: x * 255),\n",
    "    ])\n",
    ")\n",
    "\n",
    "# A dataloader that will run through all example images in one batch\n",
    "dataloader = DataLoader(example_images, batch_size=20)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_df = np.zeros((len(example_images), example_images[0][0][0].shape[0], example_images[0][0][0].shape[1]))\n",
    "for i in range(len(example_images)):\n",
    "    image_df[i,:,:] = example_images[i][0][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.cuda.is_available()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Determine whether there is a compatible GPU available\n",
    "device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Run the model on the example images\n",
    "#s2: 1st neuron; 2nd image; 3rd filter; 4th (64,64) picture;\n",
    "model = model.to(device)\n",
    "for X, y in dataloader:\n",
    "    s1, c1, s2, c2 = model.get_all_layers(X.to(device))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8, 8, 10, 400, (64, 64))"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(s2), len(s2[0]), len(s2[0][0]), len(s2[0][0][0]), s2[0][0][0][0].shape\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x187a4f64880>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWoAAAArCAYAAABRq8KeAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAjtklEQVR4nO2deZBlWV3nP79z7vbWzHy5VFZWZe29N0s3yC7Kog1oCDo4IjogGsGES4ROzDhiqIMzMeOMo6ih4zLoKLjRyCaIICAQQAP2Br13V3dVd+2V68v38m13O+fMH/dWdtHTXbRMdVd2+D4RL/Lmffe9932/e87v/H6/c9694pxjzJgxY8ZsX9SlFjBmzJgxYy7M2FGPGTNmzDZn7KjHjBkzZpszdtRjxowZs80ZO+oxY8aM2eaMHfWYMWPGbHOelKMWkdeIyGEROSIi73iqRY0ZM2bMmEeRb7aOWkQ08CDwXcAp4Fbgh51z9z318saMGTNmzJOJqF8AHHHOPeycS4Ebgdc/tbLGjBkzZsw5vCdxzC7g5Hn/nwJeeKEXTLWU27vosZZXWEtq7K+us5I1ERyesvSzgEl/xEZWJY89dk20OTOcJPIzZv0eFcmxCLHzGNiQUDKGNqSqEmoqJXEegmNoQ0QcvhgS61FXCVosifPwsKTOI5AcLcV2Yj1i6xMbjwPROqumTiA5qfPInSKUnHZapeknTHoDBAjF0TYRy8MGk9GIkfGp6Awtlhk9QIuQOoidT2x9PDE0VEIgQuIcFsEBFXHETuiYKtN6QOoUWhzLWRNPLHUdA+AQmioltoqTcYu5cBMlDh+DL5bEaRLnsxrX2VnpUpOUM9kkNZ2w0p6g0kyY9IZb3zl3igk9op3XmNRDlrMmB8MuQwcax9msiXGKSOc01ZCBjeikFSIvo6FjEuczo/us5g0CldPNKiyGGyxlTWo6YUonbJiQzGnqqji+ZyJm/R51cXSsx9AGzOo+G7aK4MicJpSMpfYU860OAH0T0vL6rOcNZv1NzqaTtPxBYQunqIjlTNZgwe+ROXBQfqahZz3Wszozfo+6MuTOkaLJnaYiGQZBcIxcgMLRyStMe31qyjKwioGN8MRgnMJXhjODCearj9o9RzGhMnrWYy2rM+0PWE3r7A438HAoERIn9Erb7Y42AEidR11SlvMGwzxgZ9glR7GSNNgTtRnZAE8MnlhOxlOIQJppLq+vcCqdwgG+Msx4fQTHUjbBtNenoSx9K6TOQ4mjoTISJ6xlDWo6YT2tkQ19ppoDQpWRO01DjzidTLEz6DKwIRWVYhEAhiakqYd0TI2mHlFXOQAjp4itj+CoqpSj/VmubKwiwMBqRjZgSg9ZM/WtflFXMZ5YNkyNKT1gw9TIraKhY9bSOgcrbVIH7byOFlv2aUPihNj6aLHE1segyJ1iwdtk5DSxC1hPa8yHm4SSMbAh62mNPVGb2PlUJMMiaBx9G9JQMZs2wiHUVUzuNB1TJXeKaa+PKc/pal4hdR6+GCb1kECK75Y4j9R5WCe0vAFJ6Usy5zEwIZHKqKmE3ClWsiYO2BV02DBVZvWAVVMjtR4zfg/nBARWswZx7rMQdYgkxyAkzsPHECnLqZM5620r36qjBvghEXkhYIAm8InHHiAibwfeDrCwS/HFT+7gXevX8/6Hrufd1/85v3z0+9lR7VHTKcf6LX5w4XZ+975XoG9u8j/+7Z/ysx/9MZqXbfB7176Pq/3CaX1mtJPfePAGXrrzYUbG5wWNR/j26lE+M7iSs+kkq2md+XCTE6MWxglvnfsyGsexbIZ5r8tnutcwsgE/MfNFbo/3cUd/D4nxuPnMXj7yvP/Nry3dwP7KGpnT/OW9L+ANV97JRz/1Il7xqjv44embUWK52h/wZ91n8we3vIIfeM7X+MTD1/DyvUcY5CG/s/j3RKI5Ywyf6l/N+08+jxfPPcKPt77MXs/jliRiwevRNhFXBSn3pwF3xHu4vnKM+5JdvKTyCO9afjVHNme5vnWSXeEGC/4G31k5wwNZjZ/+w5/ibW/9B0KVMett8tzwDHckCzySzPFHX/lO/uBV72Gf3+GjvWeTWJ8P/8ErMK/p8O+u/Cy39fZzoLLKvf0F3jz7VX7v1Kv58YUv8VuPfDefueZDfDXRTKsRP3n4zRycWOPK2hKvbdzNe9dfwkfuvo7nHzrGVDDkyOYs7738r3nrg2/mQGMdgJ+f/zQ/cvfbePnOI/zi3Jf4xGAvkcpoqpi7493c3t3LO3d9nMv9iN9sX8HNG/v4jb0f4cO955BYH+MUz689zDt//W382i/8CWfyKT6x9ix+ZuGz/F3nOn6s9RV+/pF/xVsWvsqrq6f49dWX8bMzX+JHH/hRPn71jTycQySGk3mTF4cjPj1q8Sv3fB8/f9WneWP9DG2bsmZ8DMIBL+fWZIJFr8vhbA6Aj65fx0/u+BxX+DmHM4/P96+mZyKODGZ56eRRfudTr+Xt3/VZ7u7t4vrmCc6mE7xz7qvcFE/w16sv5HWtu/mV21/Ph1/6+0wqSJ3jnnSao+kci/46e7yATRdyd7zIDzYe4DdWX8ZdG7v4tQMf5mOb13Fi1OKts19mKZ8gdj5n0in++NZv58VXHuXmm6/gMz/4Lr7v7rcyUx1Q9VL+8+6/o20j/mr9Jbxh6nZeFg24JYk4kbXomho31O+jYwM+1r2eho755Nlr6P7tAs/6N/cQKkMrGHBd9TgfXHke3zt7F1/r7+WNrVu5L95FpDJW8wYvqT7Eu07dwFt2fpWXRMv4CJ8bzaPFcjqb4rnRcb7Qv4qfa90NwE1xjS/2r+RHJm/mzXf+OHHqc9ncKr+85++YVgk/fM/b+N2rbuQ3T99APwv5uT3/yP869Ur+5tBHOZJbbtx4ATN+nxdWj3CFP+JU7nFXsou1vMmCv8HJrMWHTlzH+655D20b8IXBlfzJ+1/DG37gJr5n4g4eSBZ474kX894r/4I703muD5d4KJtg0dvk1ngP31E5zgd71zI0Id/bvJPMKW7ceCHLSYMbWvcSW58fahzjXevX8w9nruLKqRX+4/yn2Ot5fDFucDqb4r7hAon1ecvMTdwR7+W50XE0jr/pvIBrKqd4UeU4D6SzvPO+76MRJfz25e/ngXQnN1RP8O6N67llYx/vWPwEkeT4Yvn3D7+RB+/dzX+94UZeFJ2mbXy+PDpEQ434zuoxvv971p/QAT+Z0sfp0qG/wjn3XOCPy33fgHPu3c655zvnnt9seSgUG3mVeBigcawPqty/uoPDnTkePjHHqbTFaLWKTiC2AXr3kDj16dgqBsfAWSLJePH8I0x4I6wrpPplRHY6nqSuE+o6Zi7s0UmrAFRVwmreIHWa59RO8D1TdzBwAQAj49PLQ5KHmmiBhahDQ8f4YlAPV7BOyGuWB7tzpE7TlARfisi2fn/AalqnUY0JVU5FZwydw+Do2ICGHlHxMya8EZFYfNFUVcLQFqM1gC+Gs9kkARZfDMYJy6MmFS+jrhP2BWtM6z4KCDDkVWh5fWa9TaZ1n44NCMTgi0H3NSezaSYVnE6mALC+sGuiS89U+OrSXlpen4rO8MVgESyKzBZ2DMoI/YrJFbppxPF4mo6NaHoxLlE8tD7L3qjNVDhk6AQljh3hJrcv7wbgpTsf5jm1E2iEtbxJJBlVlVBVKSd7k0RisTjaeQ0ljlBgh9fFOEXfhHRMjcaJjAxN6jxmwgE1SQlVTuw0p7oTKLH4opgLNglEmK9tYnD4WHwcpmwTHVMlvWeCu4eLAGQOlDiqUkSGPVuh53xSp4mtz82n9jK0YflcxN29XfRNyHy0SVUlRGtq673rOuZVzftQKDqmygPtHQxsiChL5hTGOQIRZnUPXwxf6l3BwPnUJOVgsALAwIQcXZpl4AKuqZzi/vYOFr1NDIqBDTkULdG8J+BwexYdC6lzrByd5t4HFol0BoBBcWIwRccW7Tx1mkAMmdNAmR3FExinOH5yBpU7rFM0vRG9PKKqEmLj01Ax+6J1rFMcCFZY8Dbo5lWW8km+fnQP63m9bKsKLZaOqdI3EZnz+NDx55BRtOVADFosNWV51twZTK5Y6jfInMYgrLfrKLHMRz0ONdaoqoQHl2bJMAytz4lRi7WsTsdWiUSjxGFRzHibADzQ30m7WyMS6NgKdR1jAseUP6ChUgyK1GhsaQsoslqNI5KMzBX9raFjotJOmdOk1mPRX6emksJuYlnv1NlX+UYn2TY1VtM6K0mdzGk0lqjsxxpLTaXEThefZTRLG8V3P5W2tt5jdVRj3dbwxdKxIXHu4/WLLOUcZ9NJTqQz39QRPxlHfSuFo94jIgHwJuBjF3qB4LBYJrwRM9M9IsmphSm5UdSDhEZrUDgvgWSyeI09XqNZjTmWzpI5hwWGNuQLpw4VX2jUpG1qdK3PStqkojMMiqVkAoDcFp0rc5qNrMbJbJp/3LiaO4d7ia1P11TwxGKdEBwqGkNVpZxNi9d7V20ysgFouL51sogkzASJs1R0Rl6D2aBPbhRT/pCNtMKSCdEUKe9yNoF1QlWlZGVKGTufgfMZuADrHKfzSdbTOlWVo7FkKHZUNvHEkliPpWyCjqmROUfsfEzkOJNOcSqdZjVvEmAB6JkIp4rUf2Ad1gmhyjAVONWdYGgDIj+nndfLdNCSGs2miTClox64gNhpvrayC+sUu8MiXY9UBgKNqCgjBcowtB6XNVexTviBfXfSsz5/f/habukd3NLTMVWW8uKzv232xFZb2B+uEukcC8TlgDkwIQ01ImtolrJJhjZgJa5jEDbzCIAdjT49UyFztozCHad6RWPJUFigbepYigE9XciYCzaxWLrWJ3aaDFU4dsmxTqFxVFXCRG1E7IrOkjmPPZU2U96Qm1f20tAjRrsMVZ3Qz0L6JuJ0NoXFMqmHzFQHTOohJi8dpAg+wqppYN2j5z12HsfSGYxzbGYR1VpM5jRH4nkqfsaSqdIzEdO6T+Y8hjsdl7XWEAsW8OdG6EbGjrCHRZhUMYcaq8x73a3PWM2bW+03dh4VnbGRV1GBQeXQSSt4yrK/sopFEaicno1Yy+o8kOzkTD7FmXwKXww1lbC40GYtb2CcI3OWgQ0xruhXBsFahXVF31ZiMU4RO+Hh7gxhlKFVEYB0bID2LNYp1tIaJ4ZTZE7z7F1nADAIvjL4yqApgp2sHBjvHOxh3dTJnEJpy9DB6WyKs+kkJioGpLapklifQBsGziNzHpmDqCxzAfScx0rWZGiDR/tiHhKonLapc9dwkdgZJvSIyeaQzTxCiyPDsGkjfDFcXlthKhjSsxUyp7GuOM99E3I43knsNKumiYhjfqoYqEOVkTnH0AYsNjr4GDo2xBdDdxSR1yxNHWMcpCgmvCG7gzaBSNm7H59v6qidczmwDnwF6AEPO+fuvdBrzhk/tj5rq006tkIzSPC0ZX1Yo7fUILEeaqiI1qGhRljPsXxyaiv6BFg3dUI/J5ScyWBEZj18sdR1wmsm7yrqyV7M2bjJocZq2QgKZ11TCZP+CF/lW1GorwyRzkmPNjEOhjbAV4aNvIq5Z4JA5VRPar68fICBDZlUQ0y5KiZag3ZWY6G5yVLS5Jrm2dKAhQMwTtEZRRiESFxpB7vlEAAm9RBPFZG0LwYfSy+LCHSOQRGpovZtACUWpKjdznib50XleaG7VzRILdDJKgxNiEogjn1W0gb//fIPszdYYyHsUJOMXdUuu/wNlk8XTsc4hXWCVo7YeCTOI5KMnonQ9YxR5jPlDZiLeoRiuHV5D0MbcNPaQSIxaM8wHfSBwo6Z8wikaPhfPntga7C6vbePs8MmACtZk74JUTjapk71TExNJfhSnBeNI7EemdMcW57einqK2qWwp1EMJj1bDDIt3ccXjRJL5VjAybhVOmaLxjG0PplzRZYilszpwsG1Cz0+xXscGcyykVdxThjakHBVY51CSeGQuqbM8mzIMAvo2YggzLciLCXCpBrSsxFZGWVp3Na5b/oxSeLji6Hl9TnTnihrlIqqSoqafVs41m1hKg4F5Ker2E5AJ6uiKAbjh/szrOZNLBYtlgk94GC0go8rHFhaJVQ5QZSTNoSmH3N82KKbV5nVm2wkVfwyumzomJ6pALCcNumYKidPzHBFdHarvU7ropY74/eoSUqSeRiKtm3LoAhgmPoM+yHfNnuCno2oSc4rDj5Iq5x3GeXFd7/9/v1ohEVvSKByNrJqGf3arUF0yh+yljWoeylJu0IooMXRNyGmmROqbMtmp5ansE7omQhfCkddtI9isN/IqluOWollPamyPGwyqYZM+YMyG6yzenKKw70dZE7ho8lcURE+Opzl7GiCSLKtc1lVCaHKiVRGJIaGGpHnml4c4otlaEIMRb/diKtYVBGoSo6nLZILxim0UL4+Zp+/hnHugs74yf7g5XnOuQhYBA6IyMsfe4CIvF1EbhOR29bbDl12VBdrUqepeBnDYYhWhdj1rIatWlTm0GLxhoLuaYIyXYXCAUxEMTN+j80somsqxE5zPG5x12gP/ayYxDrRa3F/d35r5Ds9miSSrJgMEoMSS99EWxMeJnRoKdJaXwydrIpTjkEeEq06jFVYVFFiECG3j5rpvhM7seUkS01ybDkOdk2FJPNZSibIzlvxqHFMqhglgnWKXhaRlA4DYDOLWI9rNHSMwrKUFa+vSUrthJBYj6EN8cvJBwDrBDGFU1HA7rKEIxZ839D0Yt678jIypzk+mt563bqplyddMamHVFWOrw11vyhZaBwj4zPRGDJdHbCWNXi4PwNAZ7PKhDfCUxaDEAb5NwyqQxuyaSICycmNoloOVhWdkhmNKVPRqk7pZBUUFrGO1Gn6JirsYGpAEXF5vsGUkbMvhtg5vn5m93kN1zF0IZkz9EyFMphFI0RSlHXOnb9ztj434WtGhf0zCqeVW0VuNZdPFaWKcnwg0kVb3OF30cjWscYp4vVKERkiaIrPSKzPhDcqI2qfRb8NFCW3WiUp20mVPCscXEONiskpG2I1bPSq+N2irelUcH7Rtor24m2VQaBoH5HKigyoJNQ5ifUwRqiuWvp5MSiupkWJQYmjppLimDILU1hW4npRPrBwMp3eej+D0FQjuqbCpo2IggyNoFDEzie3iswpOhs17MjjSG+2dLyKu9YXME4Y5gFJXgy+5ELmLG3jc3IwxUrS2CpLpijaeZ09wRpaLGeGE6CLzLqhRlR1ij+RkFifhmRkTuMFRe33XEQ9sGExQFq/KI+JIbN6a58SR2KKgGTB76BE6OchkguBKt6ryJwGRdBgPFQ5MGVOk6IY2pDEeoSqGIwDMZhckZYZVqQyAhFGxmdtWCV1mqrkZE5R8TPEFG3FL/txpLItG1xoofSTctTOudPl3xXgIxRL9h57zHk1ao3BEamMysyQaT1gd7XDnrk2WlnCHUMOVYtOYf0iurQeWL+YKQfwgcvCZRSOrqmwu9rhqsoZqpJzXf0E3bzCfKVXrPYIEmpeSkON8MVwRX0ZJZYpb8i+YI2apFxbOYUnlqtqZ3Fh0QEy67HT77A72iCdNcXfCWFXo8ukGpY6pKhH73TsqbTZv7DGbFB8riozByWWnUGHyeqIuaCHgiJqReFLjo9FI1RVwsFqEfnPex0CsbTCARPBiBmvx75gjVmvR0NJ0QCbcFlleSs1jcQQ24Bd4QbxQsbVUTFVUNdFVGqKsiu7gzZ3r+8EYEdYlHm6WURTxfj1FIulJjmROATYXe0QqoxJlVLRGc4JL585Qsvrs6e6QSCWei1mV1CWR8QQ+vlWrW1/uMqst8kev40vhsumVxmWnnM+2GS+toktdfli2BFussvfwFQ8uqZGXccoCicyG/TxxTBRG23V6yOV4QO7Wx2sc9TKyGnB20ChaOgRWdPS9OJvaJMdWyF1RSQMRUYz73VZ2N2mKkn5XXJ2VztM+kN8KcooowVDVSXsjLos+BtFxy/nHFqVIYv+Oq1dnUczR2e26vNTXlHWa6iYad3HF+Gy6grzjR6TKqauY6JKSly284ENqKmE9OoRreaArOHQQN6w7NqzTi8Pqaqcqso4VFvdyjLORe2p87YGs+c3j3EwWmGhtclwVvGiqUeYCfsshF2m1YCrJpaJrY91qlzhkFNTKVc2lpnUQyQyRXsVQYkQSUZL9zkQrLLL2+RQa61c4ZKTOs1c0CuiwokREhk2k4iaZDRURm8UkaFQOFKji0F6xwCAFVNnsbbBodrq1rk5NwelxXF5dJYrmsuIb1EUdWSAbBAUmRGKaysn8TyLL5YFf4NQivOryvMcSNEeZvyiDU3rAdPhgKqfkqKZ1n00wr5onbkD60wFI0zZZhe9DrPeJg0/Zl99napKaKiYWZXgSzE/da4/DmxIvRZTDTOsE55TOQ7AFdVlXrnrIXZ5HXQZtMzXNjG1IjgYumIgbOd17hktouVxF3ts8U1XfYhIDVDOuV65/d3Af7nQazQO6xwKx4t3H6MqOUosnlj2NjaYrgzZ4XVRtYz+niJVdJ7Dnx/yvOgYBuhaTUON2FHdpKpSFsJOMdEjllmvx4HmCl8dXIYWS6QzpoIRQxdyUG+wO2jTVEUkPrDhVopa0Sktr8/i/lWMg8R5xM7nOdUTLO5fLZYEzjoua6yUJQyzleo5DcYpnjV1hj3hOs+KThaThgTM66JzLtS7hCrbqjXVpOiQWoqaX+YCZvweSlzZyRTWKabDYVH6kIxp3d+y4+jaEQ0dM+91ivJAOYGTOQ2BLVPAwpEpseR1aNVGnM0m+fb5o+zz19i0FaqSc3l9hdRpZqd6RaN3iswZUqNpp1WurqYYhOfXHuH03CR9E3J15TTUKCZNgox5r8Mb52/fKu0ktkiFDwQrWw5kVm9yeX0FXdptIdjgAb2jqBk7zU6/GBSMU4xmA64Iz7BqmsxGfQY2pOUNaEjGgYmig/iiOBQuE4niZTNHUVIMYjVl6eYKMEUn35mwO2ijUIQCipzI20RTTDYe9NcLp6Ag8nIm9XArSt4VdqjrmOWkSex85g6usy9YK+3kMe31iV3OnO6xI+rRsxVevftBBs6jgcG4YuJuh9/lyuDsVnszKkEjzPg9Ip0xtD41lXDF7Ao9WyEQw6y3yWrexFnYWdtkqTFNBlTm+0xGI7596qFy0sqwJyxsolB0bBVfcnb5baoCS86jm1dZCDZIjCZrwoQ3ZCltYhEMQs1LmNRDml4xCFoUqdNcHp1lWve5/sAJnhs9Or9wbkL1XEayt9reem5SD8sBtuDavWfYWemSohg4D6WKXmARtCoGwH996Ov4opj3erx84kFi67PHa2OBqspoeX06psbBoMMOf5PaxAiAhiqChFde8wAt3SdzxSSsiMM4IVJpEXlLhpZCryo/O3OarJxHKoKAQpcqe2nL67Nvos3ABKWdisUKvhj2Rm18VZS4Wl6fnvOYVOf6ZJeB86iphIlKzOUTK0yoDCgGi5bXp1r2iV4Z0DS8BCmXISqgKjlXRGe2bHqhHx8+mYh6B3CTiNwJ3AL8vXPuHy70gnNjg0X4ysn9dG3Irat7eGS1xcPdae49sZOvDfYhCqpLwlI+QeOyDsaorfS3pYrR6jsmH2Qla3Lbxt5yphkeTmd539qLyJxmQo+Yj3osxw1aus+xfIIvdC5HYdkXrLLPX8VQ1IVClXMqbXH6/h1AUfqIJOO2wX5OPThHKDkqg3ZaI5KMh9I5YufYzCPqx2Ezj/j44WuJnc8X+lfhSxE5++VMM8BGXisNq4jEMK+HDO2j4+Fa1ihWwZRliEl/xGzQo6YSVk2Do+kcQ1dMmEx/JionDEM0RXSxlE2QWQ/V9TmZFWnqkVGx7CzoQiNIuCo6w9fai9we72MlaxKIZTVtMK37rG/WinqYOHyBF+04xsqowdF4jp71+fpwL186eohOXuVkOs3Hlp9DQxmUOO4ZLfK3y9dhHLxk/hGurJwldo4PrL+Ao+kcD6XzPJTOc0dnN9XSNqfSFjvCHrO6qIufzSbp5tViDe2ZmPuSXbTzOutJlaaKOZG0WLVVbjm+t6z5WtZNnYGzfODodVjnqIopJ4+K0sS818Wthjww2rlViopLJxOJ2lp90c7rrORNHjkzw1I+UUTjKuX27h7uHyzQCooBt/tPc9w23M9fnnghq3mjmENxjtvjfXzl1D56NuLjD1/DrMrxEapKs25rDGzIvclC0d6QrSjx8HCew6tzxK5YdXHPmZ0c8Nu0dJ+erTDrbeKdiDjWaRGd9fCB0akGJzuTnE0naUhO20R8pXtwqx2dK5ucq6fGzmcpbXImnWK13WTqsOW2zX0cqKxS1zHrps6nT1xZTHCiWMont97r7uEi9ya7uf3eA9yVLGKBxFkildE2xaqHng345LGryZwlFI9OOaEHsDjZ4f7T89x08gAaR01yest1fCzPnjjNwYk1ejbixgefR+YsPpa7hos8kswWkbaUtW4bMu916JgaHzh5Hf1OhczBpCrmMj53x9VEKmNWpwRiEIGgdLwDq0icLhygSogE5vxNdvhdfByL3mZR1x5OMqlijmWzZfCkWRo0OVRbpSqGxFk6tsK6qbOUNjk+mmFofQIxWCd0bIXjo2nujnejcSzlE6z1a3xtZZGe87gl3o8viuVsgps6l9GzUZFh2Iibjh/A6ykW/CIzHTqPe0aLrOd1qqKRC0TV3/Qn5N8KItIDDl/0N774zABrl1rEk2Cs8+LxTNAIY50Xm2eCzr3OudnHe+LJ/uDln8th59zzn6L3vmiIyG1jnRePZ4LOZ4JGGOu82DxTdD4R48ucjhkzZsw2Z+yox4wZM2ab81Q56nc/Re97sRnrvLg8E3Q+EzTCWOfF5pmi83F5SiYTx4wZM2bMxWNc+hgzZsyYbc5FddTb+ZZdInJMRO4WkTtE5LZyX0tEPiMiD5V/py6Brj8VkRURuee8fY+rSwp+t7TvXSJy/SXW+asicrq06R0i8rrznvvFUudhEbnhadS5KCKfF5H7ROReEfnZcv+2sekFNG4re4pIJCK3iMidpc7/XO7fLyI3l3reX16sDREJy/+PlM/vu8Q63yMij5xnz+eW+y9ZP/qWcc5dlAeggaPAASAA7gSuvljvfxH0HQNmHrPvfwLvKLffAfz6JdD1cuB64J5vpgt4HfBJit8UvQi4+RLr/FXgPzzOsVeX5z8E9pftQj9NOncC15fbDYrbyF29nWx6AY3byp6lTerltg/cXNrob4A3lfv/CPjJcvungD8qt98EvP9pOudPpPM9wBsf5/hL1o++1cfFjKifibfsej3w3nL7vcAbnm4BzrkvAu3H7H4iXa8H/twV/BMwKSI7L6HOJ+L1wI3OucQ59whwhMe5PsxTgXPurHPua+V2D7if4i5F28amF9D4RFwSe5Y2OXdNA798OOCVwAfL/Y+15TkbfxB4lVzo53ZPvc4n4pL1o2+Vi+moH++WXRdqfE83Dvi0iNwuxd1oAHY4585d13GJ4ufy24En0rUdbfwzZfr4p+eVjraFzjL1vo4iwtqWNn2MRthm9hQRLSJ3ACvAZyii+Y4rLn/8WC1bOsvnu8A0TwOP1emcO2fP/1ba87dFJHyszpLt0I8uyL+kycSXOeeuB14L/LQ85lKtrsiJtt0SmO2qq+QPgYPAc4GzwLsuqZrzEJE68CHg55xzm+c/t11s+jgat509nXPGFXd22k0RxV95aRU9Po/VKSLXAr9IoffbgBbwC5dO4f8fF9NRn6a4XvU5dvM4t+y6VLjHv1Tr8rmUp/y7cukUfgNPpGtb2dg5t1x2EEtxi7Zz6fgl1SkiPoUD/Cvn3IfL3dvKpo+ncbvas9TWAT4PvJiiVHDu8hPna9nSWT4/QXHTkUuh8zVlick55xLgz9hG9vzncjEd9a3AZeWM8JO6ZdfThYjURKRxbpviUq33UOh7a3nYW4GPXhqF/w9PpOtjwFvKWesXAd3z0vmnncfU9b6fwqZQ6HxTuQpgP3AZxZUXnw5NAvwf4H7n3G+d99S2sekTadxu9hSRWRGZLLcrwHdR1NM/D7yxPOyxtjxn4zcCnyuzl0uh84HzBmahqKOfb89t04+eFBdzZpJiNvVBijrWLz1dM6JPQtcBilnzO4F7z2mjqJ99FngI+EegdQm0vY8izc0oamU/8US6KGapf7+0793A8y+xzr8oddxF0fh3nnf8L5U6DwOvfRp1voyirHEXcEf5eN12sukFNG4rewLPBr5e6rkH+E/l/gMUA8UR4ANAWO6Pyv+PlM8fuMQ6P1fa8x7gL3l0Zcgl60ff6mP8y8QxY8aM2eb8S5pMHDNmzJhnJGNHPWbMmDHbnLGjHjNmzJhtzthRjxkzZsw2Z+yox4wZM2abM3bUY8aMGbPNGTvqMWPGjNnmjB31mDFjxmxz/i9NZYbiIAVxcgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(c2[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('output.pkl', 'wb') as f:\n",
    "    pickle.dump(dict(s1=s1, c1=c1, s2=s2, c2=c2), f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "def image_move(image, image_param, x_coord, y_coord, path_lenght=20):\n",
    "\n",
    "     x_shifts = np.array(x_coord[1][x_coord[1] != '-']%100, dtype=int)[:path_lenght]\n",
    "     y_shifts = np.array(y_coord[1][1][y_coord[1][1] != '-']%100, dtype=int)[:path_lenght]\n",
    "\n",
    "     new_images = np.zeros((len(x_shifts),image_param+image.shape[0],image_param+image.shape[1]))\n",
    "     im_par_set = np.zeros((len(x_shifts),4))\n",
    "\n",
    "     v1 = image_param//2\n",
    "     v2 = image_param//2+image.shape[0]\n",
    "     v3 = image_param//2\n",
    "     v4 = image_param//2+image.shape[1]\n",
    "     im_par_set[0,:] = np.array([v1, v2, v3, v4])\n",
    "     new_images[0,v1:v2,v3:v4] = image\n",
    "\n",
    "     for j in range(1,new_images.shape[0]):\n",
    "          #new_images взять картинку предыдущего шага!\n",
    "\n",
    "          v1_new = int(im_par_set[j-1,0] + x_shifts[j])\n",
    "          v2_new = int(im_par_set[j-1,1] + x_shifts[j])\n",
    "          v3_new = int(im_par_set[j-1,2] + y_shifts[j])\n",
    "          v4_new = int(im_par_set[j-1,3] + y_shifts[j])\n",
    "\n",
    "          im_par_set[j,:] = np.array([v1_new, v2_new, v3_new, v4_new])\n",
    "          new_images[j,v1_new:v2_new,v3_new:v4_new] = image\n",
    "     return new_images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_images = image_move(image_df[0,:,:], 1000, x, y, 40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x187b2057df0>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(new_images[39,:,:])"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "8be52dd2d7c2878edb8d6b176d803d56e33a8f9069206cb448eb742dd5aa0267"
  },
  "kernelspec": {
   "display_name": "Python 3.10.2 64-bit",
   "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": "3.10.2"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}