simple_model_comparison.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'config' from '/home/deus/Documents/trajectory-prediction/config.py'>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from importlib import reload\n",
    "import parsing\n",
    "import model_path\n",
    "import model_goal\n",
    "import model_interface\n",
    "import visualization\n",
    "import util\n",
    "import config\n",
    "reload(parsing)\n",
    "reload(model_interface)\n",
    "reload(visualization)\n",
    "reload(util)\n",
    "reload(config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1224/1224 [00:17<00:00, 71.15it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dividing 697 human trajectories: 627 training, 70 eval, 0 test\n",
      "train (227965, 40, 2) float64 (227965, 25) float32 (227965, 40, 2) float64\n",
      "eval (21255, 40, 2) float64 (21255, 25) float32 (21255, 40, 2) float64\n",
      "(21255, 25)\n",
      "[0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
      " 0.]\n"
     ]
    }
   ],
   "source": [
    "# file_path = \"data/small.csv\"\n",
    "file_path = \"data/medium_aa.csv\"\n",
    "# file_path = \"data/atc-20121114.csv\"\n",
    "\n",
    "train_data, eval_data, test_data = parsing.parse_atc_day(file_path, train_ratio=0.9, eval_ratio=0.1)\n",
    "eval_data_x, eval_data_goals, eval_data_y = eval_data\n",
    "print(eval_data_goals.shape)\n",
    "print(eval_data_goals[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_6\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_7 (InputLayer)         [(None, 40, 2)]           0         \n",
      "_________________________________________________________________\n",
      "conv1d_6 (Conv1D)            (None, 38, 32)            224       \n",
      "_________________________________________________________________\n",
      "max_pooling1d_6 (MaxPooling1 (None, 19, 32)            0         \n",
      "_________________________________________________________________\n",
      "flatten_6 (Flatten)          (None, 608)               0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 25)                15225     \n",
      "=================================================================\n",
      "Total params: 15,449\n",
      "Trainable params: 15,449\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Epoch 1/10\n",
      "7124/7124 [==============================] - 11s 2ms/step - loss: 1.1478 - accuracy: 0.6268 - val_loss: 1.1588 - val_accuracy: 0.6095\n",
      "Epoch 2/10\n",
      "7124/7124 [==============================] - 11s 2ms/step - loss: 1.0832 - accuracy: 0.6415 - val_loss: 1.1475 - val_accuracy: 0.6104\n",
      "Epoch 3/10\n",
      "7124/7124 [==============================] - 10s 1ms/step - loss: 1.0730 - accuracy: 0.6449 - val_loss: 1.1346 - val_accuracy: 0.6182\n",
      "Epoch 4/10\n",
      "7124/7124 [==============================] - 11s 1ms/step - loss: 1.0679 - accuracy: 0.6464 - val_loss: 1.1336 - val_accuracy: 0.6231\n",
      "Epoch 5/10\n",
      "7124/7124 [==============================] - 10s 1ms/step - loss: 1.0652 - accuracy: 0.6467 - val_loss: 1.1310 - val_accuracy: 0.6223\n",
      "Epoch 6/10\n",
      "7124/7124 [==============================] - 10s 1ms/step - loss: 1.0627 - accuracy: 0.6476 - val_loss: 1.1228 - val_accuracy: 0.6230\n",
      "Epoch 7/10\n",
      "7124/7124 [==============================] - 10s 1ms/step - loss: 1.0607 - accuracy: 0.6475 - val_loss: 1.1207 - val_accuracy: 0.6285\n",
      "Epoch 8/10\n",
      "7124/7124 [==============================] - 11s 1ms/step - loss: 1.0592 - accuracy: 0.6490 - val_loss: 1.1245 - val_accuracy: 0.6231\n",
      "Epoch 9/10\n",
      "7124/7124 [==============================] - 10s 1ms/step - loss: 1.0583 - accuracy: 0.6482 - val_loss: 1.1172 - val_accuracy: 0.6267\n",
      "Epoch 10/10\n",
      "7124/7124 [==============================] - 11s 1ms/step - loss: 1.0567 - accuracy: 0.6487 - val_loss: 1.1257 - val_accuracy: 0.6204\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7fe82a44c700>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reload(model_path)\n",
    "reload(model_goal)\n",
    "\n",
    "# print(train_data[0].shape)\n",
    "# m_det = model.Model(model.simple_cnn(train_data))\n",
    "# m_det.train(train_data, eval_data)\n",
    "\n",
    "\n",
    "m_goal_estimator = model_goal.ModelGoal(model_goal.simple_cnn(train_data))\n",
    "m_goal_estimator.train(train_data, eval_data)\n",
    "\n",
    "# m_det_goal = model_path.Model(model_path.simple_cnn(train_data), uses_goal=True)\n",
    "# m_det_goal.train(train_data, eval_data)\n",
    "\n",
    "# model, _ = models.simple_lstm_goal(train_data, eval_data)\n",
    "\n",
    "# model_prob, _ = model.simple_cnn_dropout(train_data)\n",
    "\n",
    "\n",
    "# model_no_goal, _ = models.simple_lstm(train_data, eval_data)\n",
    "# model_goal, _ = models.simple_lstm_goal(train_data, eval_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_2 (InputLayer)         [(None, 40, 2)]           0         \n",
      "_________________________________________________________________\n",
      "conv1d_2 (Conv1D)            (None, 38, 32)            224       \n",
      "_________________________________________________________________\n",
      "max_pooling1d_2 (MaxPooling1 (None, 19, 32)            0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 19, 32)            0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 608)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 80)                48720     \n",
      "_________________________________________________________________\n",
      "reshape_2 (Reshape)          (None, 40, 2)             0         \n",
      "=================================================================\n",
      "Total params: 48,944\n",
      "Trainable params: 48,944\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Epoch 1/20\n",
      "7066/7066 [==============================] - 11s 2ms/step - loss: 0.0029 - mean_absolute_error: 0.0345 - val_loss: 0.0031 - val_mean_absolute_error: 0.0346\n",
      "Epoch 2/20\n",
      "7066/7066 [==============================] - 12s 2ms/step - loss: 0.0028 - mean_absolute_error: 0.0336 - val_loss: 0.0031 - val_mean_absolute_error: 0.0352\n",
      "Epoch 3/20\n",
      "7066/7066 [==============================] - 12s 2ms/step - loss: 0.0028 - mean_absolute_error: 0.0335 - val_loss: 0.0030 - val_mean_absolute_error: 0.0345\n",
      "Epoch 4/20\n",
      "7066/7066 [==============================] - 12s 2ms/step - loss: 0.0028 - mean_absolute_error: 0.0335 - val_loss: 0.0031 - val_mean_absolute_error: 0.0352\n",
      "Epoch 5/20\n",
      "7066/7066 [==============================] - 12s 2ms/step - loss: 0.0028 - mean_absolute_error: 0.0335 - val_loss: 0.0031 - val_mean_absolute_error: 0.0348\n"
     ]
    }
   ],
   "source": [
    "reload(model)\n",
    "m_prob = model.Model(model.simple_cnn(train_data))\n",
    "m_prob.train(train_data, eval_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv1d (Conv1D)              (None, 38, 32)            224       \n",
      "_________________________________________________________________\n",
      "max_pooling1d (MaxPooling1D) (None, 19, 32)            0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 608)               0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 80)                48720     \n",
      "_________________________________________________________________\n",
      "reshape (Reshape)            (None, 40, 2)             0         \n",
      "=================================================================\n",
      "Total params: 48,944\n",
      "Trainable params: 48,944\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# print(\"Deterministic model\")\n",
    "# for _ in range(5):\n",
    "#     print(model_interface.predict_once(model, eval_data_x[0])[1])\n",
    "\n",
    "# print(model.summary())\n",
    "\n",
    "# print(\"Bayesian NN\")\n",
    "# for _ in range(5):\n",
    "#     print(model_interface.predict_once(model_prob, eval_data_x[0])[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MED: 0.04966662976588962\n",
      "FDE: 0.09358999068223993\n"
     ]
    }
   ],
   "source": [
    "# eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, eval_data_goals)\n",
    "model_interface.calculate_final_metrics(model, eval_x=eval_data_x, eval_y=eval_data_y, filepath='goal_baseline')\n",
    "# model_interface.calculate_final_metrics(model, eval_x=eval_data_goal_x, eval_y=eval_data_y, filepath='goal_model_errors')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAewAAAHiCAYAAAAqIP8QAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAAAmm0lEQVR4nO3deZyVZd3H8c9PEDVQ0aBRRAELTVxyGXFLGbdcMrEytTTRUp5K08SeMn0q28xcyo1UXBLTRNRUMhWXHDUTA3PHVMQFEHADE0nTvJ4/7psaxxk4Osc552I+79drXnPOfV/nXL/zg5nv3Mu5T6SUkCRJ9W2ZWhcgSZKWzMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLVRQRKSI+Vus63ouIGFjW3b3WtbQlIm6IiBHVHtvqcdtGxGPvvTqp8xjY0vsUEc0RcUiVnuvW1qFZBultEbEwIv4eETu1esxRETEnIv4RERdGxHLVqGUJdR4UEX/+oOdpKaW0W0ppbDXHtv7DKqV0Z0pp3Y7UKX3QDGypxiJif2DZNlZdBtwHfBg4DrgyIvqWj9kFOAbYERgArA38qFMK7oB63YqXcmBgq0uLiKcj4nsRMTUi5kXEbyJi+XLdKhFxXUS8UK67LiL6l+t+BmwLnBURCyLirBZPu1NEPBER8yNidETEYuZfGfgh8J1Wy9cBNgV+mFL6Z0rpKuAh4PPlkBHABSmlR1JK84CfAAdV+Jq7RcQpEfFiREwHPt26poi4ICJmR8SsiPhp+Zj1gHOArcrXPL8cv1z5fM9GxNyIOCciVijXNUXEzIj4bkTMAX4TEcdHxBURcUlEvBoRD0XEOuW/w/MRMSMiPtWinv/syVi0hV/ONy8inoqI3doZ+7GIuD0iXilf6+Xl8jvK4Q+Ur2PfRXW2eJ41I+L35b/9S63+faWaMLAl2B/YBfgosA7wf+XyZYDfUGzBrgX8EzgLIKV0HHAncHhKqVdK6fAWz7cHsDmwEbBP+dztOQE4G5jTavn6wPSU0qstlj1QLl+0/oFW6xoi4sNLerHAoWWNmwCNwN6t1l8EvAV8rBzzKeCQlNKjwNeAu8vX3LscfyJF3zYuH7MG8IMWz7casCpFH0eWyz4D/BZYhWIvwkSKfq8B/Bg4dzH1bwE8BvQBTgIuaOePop8AN5Vz9AfOBEgpbVeu/0T5Oi5v+aCI6AZcBzwDDCxrGreYeqROYWBLcFZKaUZK6WXgZ8AXAVJKL6WUrkopLSyD82fAsAqe78SU0vyU0rPAbRRB9i4R0QhsQxkkrfQCXmm17BVgxXbWL7q9Iku2D3Bai9f88xY1NQC7A99KKb2WUnoe+BWwXzuvIShC+KiU0stln05oNf5tij0Fb6SU/lkuuzOlNDGl9BZwBdCXom9vUoTjwIjo3U79z6SUzksp/RsYC6wONLQx7k2KPxL6pZReTylVeux9KNAP+N+yB+/lsdIHxuNJEsxocfsZil/WRMSHKMJqV4qtNIAVI6JbGRbtabm1vJAiXN8hIpYBfg0cmVJ6q40NxAXASq2WrQS82s76RbdfZcn68e7XvMgAiuPps1vUtEyr8S31BT4E3NtifADdWox5IaX0eqvHzW1x+5/Aiy16uijUewHz25jzP/1NKS0s531XjykOM/wE+GtEzANOTSld2M7raGlNij8K3qpgrNRpDGyp+AW9yFrAc+Xto4F1gS1SSnMiYmOK3beLkqkjH3W3EsXu6MvLwFkUcDMj4gvAI8DaEbFii93inwB+V95+pLw/vsW6uSmllyqYezbvfs2LzADeAPq0E1itX/OLFAG7fkppVjvz1eQjAVNKcyh2/xMRnwRuiYg7UkrTlvDQGcBaEdHd0FY9cZe4BIdFRP+IWJXibOxFxzRXpAij+eW6H7Z63FyKs7Pfj1cotnQ3Lr92L5dvBtyTUnocuB/4YUQsHxGfpTgmflU57mLgqxExpNx1/H8Ux56B/5x8dXw7c48Hjihf8yoUZ5sDkFKaTXHc99SIWCkilomIj0bEokMBc4H+EdGjHP82cB7wq4j4SDn3GuVZ7DUVEV9YdJIgMI/iD4e3y/uL+7f7K8UfNSdGRM+y/9t8sNVKS2ZgS8VW603AdOBJ4Kfl8tOAFSi2IicBN7Z63OnA3uXZyme8lwlTYc6iL+CFctXclNK/ytv7UWyFz6M4sWvvlNIL5eNvpDjh6jbgWYrd2i3/oFgTuKud6c+jOMnrAeBvwO9brT8Q6AFMLee+kuI4McCfKLbu50TEi+Wy7wLTgEkR8Q/gFoo9E7W2OXBPRCwAJlAcfpherjseGFueyb9PyweVu+Y/Q3EC3bPATGDfTqtaakekVJO9VVJdiIinKc6AvqXWtVRLuVU5PqW0da1rkVQ9HsOWljIppZmAYS0tZdwlLklSBtwlLklSBtzCliQpAwa2JEkZqNuTzvr06ZMGDhxY6zLe5bXXXqNnz561LqPu2afK2KfK2avK2KfK1WOv7r333hdTSn3bWle3gT1w4ECmTJlS6zLepbm5maamplqXUffsU2XsU+XsVWXsU+XqsVcR8Ux769wlLklSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpA1UJ7Ii4MCKej4iH21kfEXFGREyLiAcjYtNqzCtJUldRrS3si4BdF7N+N2Bw+TUSOLtK80qS1CVUJbBTSncALy9myHDg4lSYBPSOiNWrMbckSV1BZx3DXgOY0eL+zHKZJEmqQPdaF9BSRIyk2GVOQ0MDzc3NtS2oDQsWLKjLuuqNfaqMfaqcvaqMfapcbr3qrMCeBazZ4n7/ctk7pJTGAGMAGhsbU1NTU6cU9140NzdTj3XVG/tUGftUOXtVGftUudx61Vm7xCcAB5Zni28JvJJSmt1Jc0uSlL2qbGFHxGVAE9AnImYCPwSWBUgpnQNcD+wOTAMWAgdXY15JkrqKqgR2SumLS1ifgMOqMZckSV2RVzqTJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGWgKoEdEbtGxGMRMS0ijmlnzD4RMTUiHomI31VjXkmSuoruHX2CiOgGjAZ2BmYCkyNiQkppaosxg4HvAduklOZFxEc6Oq8kSV1JNbawhwLTUkrTU0r/AsYBw1uNORQYnVKaB5BSer4K80qS1GVUI7DXAGa0uD+zXNbSOsA6EXFXREyKiF2rMK8kSV1Gh3eJv4d5BgNNQH/gjojYMKU0v+WgiBgJjARoaGigubm5k8qr3IIFC+qyrnpjnypjnypnrypjnyqXW6+qEdizgDVb3O9fLmtpJnBPSulN4KmIeJwiwCe3HJRSGgOMAWhsbExNTU1VKK+6mpubqce66o19qox9qpy9qox9qlxuvarGLvHJwOCIGBQRPYD9gAmtxlxDsXVNRPSh2EU+vQpzS5LUJXQ4sFNKbwGHAxOBR4HxKaVHIuLHEbFnOWwi8FJETAVuA/43pfRSR+eWJKmrqMox7JTS9cD1rZb9oMXtBIwqvyRJ0nvklc4kScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDVQnsiNg1Ih6LiGkRcUwb60dFxNSIeDAibo2IAdWYV5KkrqLDgR0R3YDRwG7AEOCLETGk1bD7gMaU0kbAlcBJHZ1XkqSupBpb2EOBaSml6SmlfwHjgOEtB6SUbkspLSzvTgL6V2FeSZK6jGoE9hrAjBb3Z5bL2vNV4IYqzCtJUpfRvTMni4gDgEZgWDvrRwIjARoaGmhubu684iq0YMGCuqyr3tinytinytmrytinyuXWq2oE9ixgzRb3+5fL3iEidgKOA4allN5o64lSSmOAMQCNjY2pqampCuVVV3NzM/VYV72xT5WxT5WzV5WxT5XLrVfV2CU+GRgcEYMiogewHzCh5YCI2AQ4F9gzpfR8FeaUJKlL6XBgp5TeAg4HJgKPAuNTSo9ExI8jYs9y2MlAL+CKiLg/Iia083RSVTz0EKRU6yokqXqqcgw7pXQ9cH2rZT9ocXunaswjVeLxx2GTTWDrreEXv4CttoJ58+Dqq+H+++FDH4KVVoKVV4Y+faChATbfHHr2rHXlktS+Tj3pTOoMgwbBWWfB8ccXob3VVnDfffD669CrF7z5JrzR6iyKj38cbrwRBnhJH0l1ykuTaqmz7LLwta/BtGnw4x/Dyy/DwQfDlCnwj38Uwf366zBnDjz8MFx+eXF7662LXemSVI8MbC21evWC738f/v53+PWvYbPNIKJYt9xyxa7w9deHffaBO+8s1m2zDVx1VW3rlqS2GNgSsMEGcPfdMGQI7L03/OY3ta5Ikt7JwJZKa64Jd9wBO+wA3/xmsUtdkuqFgS210KMHjB1bHAc/4IDiBDVJqgcGttRK//5wzjlwzz1w1FG1rkaSCr6tS2rDvvsWZ5Wfcgqstx4cdlitK5LU1RnYUjtOPLE4w/zII4v3bR911H/PMpekzuYucakd3brBZZfBnnvC0UfDuusWx7XPOQeefbbW1UnqagxsaTF69Srel33KKfDkk3DppfD1rxdXRBsypNjqvvder1su6YNnYEtLEFFsYT/8MMydC48+CqeeWoT22WdDY2PxlrC11oKPfQx++tPi2uWSVE0GtlSh9daDj3ykuO74qFFwww0we3axi7ypCXbcEdZeu7i62lprFd8XLKh11ZKWFga21AGrrAL/8z9wySXF1dFuugkefBA+/eliS3uddYrlb79d60ol5c7Alqpsww1h3LjiUqcDBsBXvlJco/zpp2tdmaScGdjSB2TLLeEvf4Hf/hamTi0+o3vChFpXJSlXBrb0AYoo3gr2t78Vx7eHD4dvf9tLnkp67wxsqRN89KNw113wjW8UZ5hvsQX89a++HUxS5QxsqZMsvzyMHg1XXw0zZhShvcMOTTz5ZK0rk5QDA1vqZHvtVbyXe5ENNoBvfasIcUlqj4Et1UCfPsXu8PHj/8K++xZb3muvDQcfXLy/e/78Wlcoqd744R9SDfXt+y8uugh+9KPi2Pb558NFFxUnq33iE/CpT8FOO8FrrxXv7371Vfjyl2GjjWpduaTOZmBLdWDAADjjDPj5z4uT0f78Z/jTn+BXv4KTTirGRED37sV1zYcNgyOOKM4679attrVL6hzuEpfqSM+esP32xWVNb7sNXnoJbrwR7rmn2LqeMwdOPhmeeQY+//niA0jOOqu4xrmkpZuBLdWxFVeEXXaBoUOLMF911eJ93NOmwfjxxaeJffObxfHvs8767yVQX3ut+JK09DCwpQx16wZf+ELx0Z4PPQTbbVcEd0NDEeq9ehVf++1XbJlLyp/HsKXMbbABXH998Vndzc2w3HLFx32+8AKcdhpMnlxcuGXePBg8GPbZB3bfHXr0qHXlkt4LA1taCiy6BOoBB7xz+Wc+A8ccU2xlr7pq8Wlil10Gm28Of/wj9O1bm3olvXfuEpeWYk1NMGlS8clhEycWn9996aXFbvShQ+H22985/tJLYeBAWGaZ4vull9agaEltcgtb6kKWXRa+9KViF/n++xeBPmwYHHJI8YEkhx8OCxcWY595BkaOLG7vv3/NSpZUcgtb6oK22AIeeKB43/fMmcXFWL761f+G9SILF8Jxx9WmRknvZGBLXVTPnsXx7ccfLy7S0t4nhz37TCoGSaopA1vq4pZZprhYy4ABba9PKz/DPt//ONeN2oO37ry9OP38hReKTfMnn4THHoN//atzi5a6II9hSwLgZz8rjlm33C2+/PL/Zrsv/ZHb+i7HFcv8kUFX/ZFv/wUOvg9WeKvFgz/0Idh6a9h22+L7FlsUV32RVDVuYUsCihPLxowptrQjiu/nn9+Nib8+jFk/eIUrdxrDR/qvy2GfhjW+/yG+dfKOTD3vBBg7tjgA/vzzcPzxsPPO0Ls37LAD3HVXrV+WtNQwsCX9x/77w9NPF5c4ffrp/54d3qNbDz6/zaHcffSjNI9o5lNDPsOv/3kH6886lq3fPIfzD9qI+ffcXlyd5cYb4dhjiw/9/uQnizeDP/RQLV+WtFQwsCVVLCIYNnAY4/Yex6xRszj1U6cy//X5HPqHQ2k4pYHhNxzI+X1n8Nx3vl5c8PyEE+DOO2HjjeFrXyuOd0t6XzyGLel96duzL6O2GsVRWx7FlOemcOlDl/L7R3/PhMcmALBRw0bs2rgru9x+EducN5Hlzj0fzj232FU+YgRsuWWx371bt+LMt4jiS1KbDGxJHRIRbL7G5my+xub8apdf8fDzD3PDtBu4YdoN/HLSLznp7bfo2a8ne120Jwc+tRI7Xngb3UaMaPvJVl652BI/6qjik0wk/YeBLalqIoINGzZkw4YN+c423+HVN17l9mdu5w+P/YHxU8dz6b/ns8bha/DlhoP4xsINWPP514s3gL/9dvE1dWrxgd9nnw1nnll8Usnyy9f6ZUl1wcCW9IFZcbkV2WOdPdhjnT04fbfTue7x6xj7wFhOeuJiTibYZ/19OHKLI9mi/xb/fdBjjxWfCzpiBIwaBQceCIceWrsXIdUJTzqT1CmW7748ew/Zmz988Q9MP2I6R25xJNc9fh1bXrAlW5y/Bb+57zcsfHMhrLtu8UHfN99cHO8+80wYMoRNv/51uP/+Wr8MqWYMbEmdbkDvAZy6y6nMGjWLM3Y9g1def4WvTPgKq5+6Ol+/7utMnn0vaccdYfx4mDULfvlLVpg1C37yk1qXLtWMgS2pZlZcbkW+ucU3efSwR7n9oNsZvu5wLnrgIoaeP5SPj/44P7/z57yy0nJw1FG8MGwYXHst7LEH/O538NprtS5f6lQGtqSaiwi2G7AdF3/2YmYfPZvzPnMe/Vbsx7F/OpZBpw/iuFuP409f3hWOPrr4mLH994f+/eHUU+GNN2pdvtQpDGxJdaX38r05ZNNDuG3Ebdw78l62HbAtJ951Il9+9HA2HjSRE3/3DZ65cVzxPu5vfxsGDy6Oc//zn7UuXfpAGdiS6tamq2/Ktftdy6xRszjso4exwrIr8L0/HcugSV9k+Ige3DD+BN4cuCYccQQMHFh8Xujf/tb+Z4VKGTOwJdW91Xqtxt799+bur97N9COmc+y2x3L3jLvZfeqx9PvM4xx23l78rWndYhf5ZpsVZ5r/4hfwyiu1Ll2qGgNbUlYGrTKIn+7wU2YcNYNr9r2GHQftyIVzb2SzIXey2elDOPu0/Zm/Zt9ia3vnnWHBglqXLFWFgS0pS8t1X47hHx/OuL3H8dyo5zhztzP59zLBN+ZfSr8d7mPE6dtz5wv3knbfrXhrmJQ5A1tS9lZZYRUOH3o49/3PfUw+dDIjPjGCq1+bwnYHvc2Qje/i5wd9lKf/73CYO7fWpUrvm4EtaakRETT2a+TsPc5m9tGzuXDPC/nwuptw7CffYNCyo9nq+6tz2jHDeH6mH/Op/BjYkpZKPXv05OBNDubPh93LU0c+xc8/cTSv912Fo1a4gzXGfJzP/WwjbnviJpJnlCsTBrakpd7A3gM5Zq9TuO9nL/HIztfwrbkDufOVh9jhd7uw/qlr86PmH/H3F/9e6zKlxTKwJXUpQ7YezsnnPsWMba7kgjtWoc/Up/lR8/GsN3o9NhuzGWfccwbzX59f6zKldzGwJXVJyw//PF+59lnu6Hk4s06F0+7syTJPP8uRNx7JGietxiFXfJlJMye5y1x1w8CW1HX16gVnnsnqf7iNIxuGM/nXb3LvufClKW8w7v5L2OqCrdho9Aacec+ZbnWr5gxsSWpqgksvhRdfZNMpszjvu39m9rThjJkAy099nCNuPIJ+p/bjkAmHcP+c+2tdrbooA1uSFuneHfr1g222YcXLr+HQMVOY/PdtufdcOODl/lz2wCVscu4mDLtoGDc96Rnm6lwGtiS1Z7PN4NZb2fSA/2XMr6Yx64Q3OOXmZXjq2QfY5ZJd2O6i7bhq6lW8+e83a12pugADW5IWJwJOOgnmzaP3jc0cvdpneeLHrzD6iXV49sUn2fuKvRlw2gCOueUYHnvRC7Log2NgS1IlVl4Zhg2DK65gubPH8I3r5jL9e3P4w8u70th7CKf85RTWG70ez736XK0r1VLKwJak9yICDj0Upk2j2zcOZ49f38KEQ25l5tWDuGThrvS7/8laV6illIEtSe9Hnz5wxhkwbRqcfjqrfWRtvnTarbDddnDxxbWuTkuhDgV2RKwaETdHxBPl91XaGLNxRNwdEY9ExIMRsW9H5pSkujJgABxxBEycCC+9BFtvDUcfXdyWqqijW9jHALemlAYDt5b3W1sIHJhSWh/YFTgtInp3cF5Jqj+9esE558D8+fDd79a6Gi1lOhrYw4Gx5e2xwF6tB6SUHk8pPVHefg54HujbwXklqT5tuCGMGgUvvghvvVXrarQU6d7BxzeklGaXt+cADYsbHBFDgR6AZ2VIWnqdcAJ061brKrSUiSVdqScibgFWa2PVccDYlFLvFmPnpZTedRy7XLc60AyMSClNamfMSGAkQENDw2bjxo2r4CV0rgULFtCrV69al1H37FNl7FPl7FVl7FPl6rFX22+//b0ppca21i0xsBcnIh4DmlJKsxcFckpp3TbGrUQR1ieklK6s5LkbGxvTlClT3ndtH5Tm5maamppqXUbds0+VsU+Vs1eVsU+Vq8deRUS7gd3RY9gTgBHl7RHAtW1M3gO4Gri40rCWJEnv1NHAPhHYOSKeAHYq7xMRjRFxfjlmH2A74KCIuL/82riD80qS1KV06KSzlNJLwI5tLJ8CHFLevgS4pCPzSJLU1XmlM0mSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJykCHAjsiVo2ImyPiifL7Ku2MWysiboqIRyNiakQM7Mi8kiR1NR3dwj4GuDWlNBi4tbzflouBk1NK6wFDgec7OK8kSV1KRwN7ODC2vD0W2Kv1gIgYAnRPKd0MkFJakFJa2MF5JUnqUjoa2A0ppdnl7TlAQxtj1gHmR8TvI+K+iDg5Irp1cF5JkrqUSCktfkDELcBqbaw6DhibUurdYuy8lNI7jmNHxN7ABcAmwLPA5cD1KaUL2phrJDASoKGhYbNx48a9pxfTGRYsWECvXr1qXUbds0+VsU+Vs1eVsU+Vq8debb/99vemlBrbWtd9SQ9OKe3U3rqImBsRq6eUZkfE6rR9bHomcH9KaXr5mGuALSlCvPVcY4AxAI2NjampqWlJ5XW65uZm6rGuemOfKmOfKmevKmOfKpdbrzq6S3wCMKK8PQK4to0xk4HeEdG3vL8DMLWD80qS1KV0NLBPBHaOiCeAncr7RERjRJwPkFL6N/Bt4NaIeAgI4LwOzitJUpeyxF3ii5NSegnYsY3lU4BDWty/GdioI3NJktSVeaUzSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGOhTYEfGFiHgkIt6OiMbFjNs1Ih6LiGkRcUxH5pQkqSvq6Bb2w8DngDvaGxAR3YDRwG7AEOCLETGkg/NKktSldO/Ig1NKjwJExOKGDQWmpZSml2PHAcOBqR2ZW5KkrqQzjmGvAcxocX9muUySJFVoiVvYEXELsFobq45LKV1bzWIiYiQwEqChoYHm5uZqPn1VLFiwoC7rqjf2qTL2qXL2qjL2qXK59WqJgZ1S2qmDc8wC1mxxv3+5rK25xgBjABobG1NTU1MHp66+5uZm6rGuemOfKmOfKmevKmOfKpdbrzpjl/hkYHBEDIqIHsB+wIROmFeSpKVGR9/W9dmImAlsBfwxIiaWy/tFxPUAKaW3gMOBicCjwPiU0iMdK1uSpK6lo2eJXw1c3cby54DdW9y/Hri+I3NJktSVeaUzSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScqAgS1JUgYMbEmSMmBgS5KUAQNbkqQMGNiSJGXAwJYkKQMGtiRJGTCwJUnKgIEtSVIGDGxJkjJgYEuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIyYGBLkpQBA1uSpAwY2JIkZcDAliQpAwa2JEkZMLAlScpApJRqXUObIuIF4Jla19GGPsCLtS4iA/apMvapcvaqMvapcvXYqwEppb5trajbwK5XETElpdRY6zrqnX2qjH2qnL2qjH2qXG69cpe4JEkZMLAlScqAgf3ejal1AZmwT5WxT5WzV5WxT5XLqlcew5YkKQNuYUuSlAEDewki4gsR8UhEvB0R7Z5NGBG7RsRjETEtIo7pzBrrQUSsGhE3R8QT5fdV2hm3VkTcFBGPRsTUiBjYyaXWVCV9ioiNI+Lu8v/dgxGxby1qrZUl/SxFxKjy/86DEXFrRAyoRZ21VsnvnIjYp+zVIxHxu86usR5ExIUR8XxEPNzO+oiIM8o+PhgRm3Z2jZUysJfsYeBzwB3tDYiIbsBoYDdgCPDFiBjSOeXVjWOAW1NKg4Fby/ttuRg4OaW0HjAUeL6T6qsXlfRpIXBgSml9YFfgtIjo3Xkl1k6FP0v3AY0ppY2AK4GTOrfK2qukTxExGPgesE35f+lbnV1nnbiI4ueoPbsBg8uvkcDZnVDT+2JgL0FK6dGU0mNLGDYUmJZSmp5S+hcwDhj+wVdXV4YDY8vbY4G9Wg8of6F0TyndDJBSWpBSWthpFdaHJfYppfR4SumJ8vZzFH/UtHkhhaXQEn+WUkq3tfh/Mwno38k11oNKfuccCoxOKc0DSCl1tT+OAUgp3QG8vJghw4GLU2ES0DsiVu+c6t4bA7s61gBmtLg/s1zWlTSklGaXt+cADW2MWQeYHxG/j4j7IuLkckuhK6mkT/8REUOBHsCTH3RhdeK9/ix9FbjhA62oPlXSp3WAdSLiroiYFBGL28rsyrL5/d291gXUg4i4BVitjVXHpZSu7ex66tXi+tTyTkopRURbbz/oDmwLbAI8C1wOHARcUN1Ka6sKfVr0PKsDvwVGpJTerm6V+YuIA4BGYFita6lT3Sl28zZR7IW4IyI2TCnNr2VRev8MbCCltFMHn2IWsGaL+/3LZUuVxfUpIuZGxOoppdll0LS1+20mcH9KaXr5mGuALVnKArsKfSIiVgL+SPFH46QPqNR6VNHPUkTsRPEH0LCU0hudVFs9qaRPM4F7UkpvAk9FxOMUAT65c0rMRja/v90lXh2TgcERMSgiegD7ARNqXFNnmwCMKG+PANraMzGZ4vjQouOxOwBTO6G2erLEPpX/h66mOK52ZSfWVg+W+LMUEZsA5wJ7dtXjslT2O+caiq1rIqIPxS7y6Z1YYy4mAAeWZ4tvCbzS4rBVfUkp+bWYL+CzFH+pvgHMBSaWy/sB17cYtzvwOMWxxuNqXXcN+vRhirOenwBuAVYtlzcC57cYtzPwIPAQxdmbPWpde731CTgAeBO4v8XXxrWuvRN79K6fJeDHFAFN2be5LXozodY112mfAvglxR/FDwH71brmGvXpMmB2+TM1k+K8h68BX2vRp9FlHx+ieAdCzetu68srnUmSlAF3iUuSlAEDW5KkDBjYkiRlwMCWJCkDBrYkSRkwsCVJyoCBLUlSBgxsSZIy8P9h/qgHurt5gQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "N = 0\n",
    "multiplier = 200\n",
    "amount = 3\n",
    "\n",
    "for i in range(amount):\n",
    "    n = N + multiplier * i\n",
    "\n",
    "    x = eval_data_x[n]\n",
    "    y = eval_data_y[n]\n",
    "\n",
    "    predictions = [model_interface.predict_once(model, x)]\n",
    "    visualization.draw_path(eval_data_x[n], eval_data_y[n], predictions=predictions, name=f\"path {n}, deterministic\")\n",
    "    predictions = model_interface.prediction_sampling(model_prob, x, samples=5)\n",
    "    visualization.draw_path(eval_data_x[n], eval_data_y[n], predictions=predictions, name=f\"path {n}, non-deterministic\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Goal 2, 0: 1.00\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Goal 2, 0: 0.90\n",
      "Goal 3, 1: 0.10\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Goal 2, 0: 0.70\n",
      "Goal 1, 0: 0.10\n",
      "Goal 2, 1: 0.10\n",
      "Goal 3, 1: 0.10\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Goal 2, 0: 0.40\n",
      "Goal 1, 0: 0.20\n",
      "Goal 1, 1: 0.20\n",
      "Goal 2, 1: 0.10\n",
      "Goal 3, 1: 0.10\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Goal 3, 1: 0.40\n",
      "Goal 3, 0: 0.20\n",
      "Goal 2, 1: 0.20\n",
      "Goal 1, 0: 0.10\n",
      "Goal 2, 0: 0.10\n",
      "Goal 1, 1: 0.10\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare prediction efficency with goals with various noise levels\n",
    "N = 1000\n",
    "\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, eval_data_goals)\n",
    "visualization.draw_path(eval_data_x[N], eval_data_y[N], eval_data_goals[N], prediction_model=model, name=\"accurate goal\")\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.3)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "visualization.draw_path(eval_data_x[N], eval_data_y[N], inaccurate_goals[N], prediction_model=model, name=\"0.3 noise goal\")\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.5)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "visualization.draw_path(eval_data_x[N], eval_data_y[N], inaccurate_goals[N], prediction_model=model, name=\"0.5 noise goal\")\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.7)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "visualization.draw_path(eval_data_x[N], eval_data_y[N], inaccurate_goals[N], prediction_model=model, name=\"0.7 noise goal\")\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.9)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "visualization.draw_path(eval_data_x[N], eval_data_y[N], inaccurate_goals[N], prediction_model=model, name=\"0.9 noise goal\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MED: 0.04022140546204346\n",
      "FDE: 0.06597747260414283\n",
      "MED: 0.04192719297606437\n",
      "FDE: 0.07061376601379137\n",
      "MED: 0.043691483300276675\n",
      "FDE: 0.07511971939235268\n",
      "MED: 0.04592443480722182\n",
      "FDE: 0.08082917048715287\n",
      "MED: 0.047317980607863096\n",
      "FDE: 0.0840148221983325\n"
     ]
    }
   ],
   "source": [
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, eval_data_goals)\n",
    "model_interface.calculate_final_metrics(model, eval_x=eval_data_goal_x, eval_y=eval_data_y, filepath='goal_noise_0_0')\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.3)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "model_interface.calculate_final_metrics(model, eval_x=eval_data_goal_x, eval_y=eval_data_y, filepath='goal_noise_0_3')\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.5)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "model_interface.calculate_final_metrics(model, eval_x=eval_data_goal_x, eval_y=eval_data_y, filepath='goal_noise_0_5')\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.7)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "model_interface.calculate_final_metrics(model, eval_x=eval_data_goal_x, eval_y=eval_data_y, filepath='goal_noise_0_7')\n",
    "\n",
    "inaccurate_goals = util.apply_randomness_to_adjacent_goal_predictions(eval_data_goals, amount=0.9)\n",
    "eval_data_goal_x = model_interface.concatenate_x_goal(eval_data_x, inaccurate_goals)\n",
    "model_interface.calculate_final_metrics(model, eval_x=eval_data_goal_x, eval_y=eval_data_y, filepath='goal_noise_0_9')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9 4 1\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "N = 0\n",
    "visualization.draw_path(eval_data_x[N], eval_data_y[N], prediction_model=model_no_goal, name=\"no goal prediction\")\n",
    "visualization.draw_path(eval_data_x[N], eval_data_y[N], eval_data_goals[N], prediction_model=model_goal, name=\"with accurate goal prediction\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "e7370f93d1d0cde622a1f8e1c04877d8463912d04d973331ad4851f04de6915a"
  },
  "kernelspec": {
   "display_name": "Python 3.8.10 64-bit",
   "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": "3.8.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}