29-TensorflowGiris.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataFrame = pd.read_excel(\"bisiklet_fiyatlari.xlsx\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Fiyat</th>\n",
       "      <th>BisikletOzellik1</th>\n",
       "      <th>BisikletOzellik2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>807.673876</td>\n",
       "      <td>1749.628226</td>\n",
       "      <td>1749.590668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>959.227520</td>\n",
       "      <td>1748.007826</td>\n",
       "      <td>1751.824206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>718.020033</td>\n",
       "      <td>1750.122967</td>\n",
       "      <td>1747.977026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>945.668885</td>\n",
       "      <td>1749.916440</td>\n",
       "      <td>1750.771646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>955.542968</td>\n",
       "      <td>1750.780519</td>\n",
       "      <td>1750.592430</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Fiyat  BisikletOzellik1  BisikletOzellik2\n",
       "0  807.673876       1749.628226       1749.590668\n",
       "1  959.227520       1748.007826       1751.824206\n",
       "2  718.020033       1750.122967       1747.977026\n",
       "3  945.668885       1749.916440       1750.771646\n",
       "4  955.542968       1750.780519       1750.592430"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataFrame.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sbn\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x7ffba1547390>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 540x540 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sbn.pairplot(dataFrame)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## veriyi test/train olarak ikiye ayırmak"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# y = wx + b\n",
    "# y -> label\n",
    "y = dataFrame[\"Fiyat\"].values\n",
    "\n",
    "# x -> feature (özellik)\n",
    "x = dataFrame[[\"BisikletOzellik1\",\"BisikletOzellik2\"]].values\n",
    "\n",
    "x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.33,random_state=15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(670, 2)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(330, 2)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(670,)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(330,)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MinMaxScaler(copy=True, feature_range=(0, 1))"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler.fit(x_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = scaler.transform(x_train)\n",
    "x_test = scaler.transform(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.3177906 , 0.64341466],\n",
       "       [0.61991638, 0.89583174],\n",
       "       [0.53950097, 0.0980286 ],\n",
       "       ...,\n",
       "       [0.2352117 , 0.52644765],\n",
       "       [0.7576794 , 0.19157421],\n",
       "       [0.4292982 , 0.16530301]])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Dense(4,activation=\"relu\"))\n",
    "model.add(Dense(4,activation=\"relu\"))\n",
    "model.add(Dense(4,activation=\"relu\"))\n",
    "\n",
    "model.add(Dense(1))\n",
    "\n",
    "model.compile(optimizer = \"rmsprop\",loss = \"mse\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 670 samples\n",
      "Epoch 1/250\n",
      "670/670 [==============================] - 1s 1ms/sample - loss: 795895.9134\n",
      "Epoch 2/250\n",
      "670/670 [==============================] - 0s 50us/sample - loss: 795754.8935\n",
      "Epoch 3/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 795628.2709\n",
      "Epoch 4/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 795503.7696\n",
      "Epoch 5/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 795366.9789\n",
      "Epoch 6/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 795216.5517\n",
      "Epoch 7/250\n",
      "670/670 [==============================] - 0s 48us/sample - loss: 795051.8409\n",
      "Epoch 8/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 794872.1590\n",
      "Epoch 9/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 794675.4231\n",
      "Epoch 10/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 794459.8364\n",
      "Epoch 11/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 794225.2847\n",
      "Epoch 12/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 793966.5489\n",
      "Epoch 13/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 793680.8845\n",
      "Epoch 14/250\n",
      "670/670 [==============================] - 0s 48us/sample - loss: 793364.7595\n",
      "Epoch 15/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 793015.5470\n",
      "Epoch 16/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 792630.6049\n",
      "Epoch 17/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 792210.4789\n",
      "Epoch 18/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 791754.9953\n",
      "Epoch 19/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 791263.5659\n",
      "Epoch 20/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 790734.0006\n",
      "Epoch 21/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 790163.4550\n",
      "Epoch 22/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 789549.6032\n",
      "Epoch 23/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 788890.6651\n",
      "Epoch 24/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 788185.5655\n",
      "Epoch 25/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 787431.7062\n",
      "Epoch 26/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 786624.2666\n",
      "Epoch 27/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 785764.2147\n",
      "Epoch 28/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 784848.3674\n",
      "Epoch 29/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 783874.3144\n",
      "Epoch 30/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 782838.6285\n",
      "Epoch 31/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 781745.7407\n",
      "Epoch 32/250\n",
      "670/670 [==============================] - 0s 48us/sample - loss: 780584.9166\n",
      "Epoch 33/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 779346.8907\n",
      "Epoch 34/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 778046.9670\n",
      "Epoch 35/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 776667.1578\n",
      "Epoch 36/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 775219.5116\n",
      "Epoch 37/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 773690.7711\n",
      "Epoch 38/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 772083.5007\n",
      "Epoch 39/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 770383.1340\n",
      "Epoch 40/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 768602.4677\n",
      "Epoch 41/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 766734.8317\n",
      "Epoch 42/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 764770.1748\n",
      "Epoch 43/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 762717.4022\n",
      "Epoch 44/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 760560.5549\n",
      "Epoch 45/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 758308.5319\n",
      "Epoch 46/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 755953.6282\n",
      "Epoch 47/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 753508.0636\n",
      "Epoch 48/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 750924.4687\n",
      "Epoch 49/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 748242.4683\n",
      "Epoch 50/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 745451.5519\n",
      "Epoch 51/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 742538.9534\n",
      "Epoch 52/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 739499.2267\n",
      "Epoch 53/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 736339.0597\n",
      "Epoch 54/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 733068.2688\n",
      "Epoch 55/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 729665.8056\n",
      "Epoch 56/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 726139.1172\n",
      "Epoch 57/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 722450.4168\n",
      "Epoch 58/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 718646.5121\n",
      "Epoch 59/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 714712.6244\n",
      "Epoch 60/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 710622.4905\n",
      "Epoch 61/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 706399.8692\n",
      "Epoch 62/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 702032.5767\n",
      "Epoch 63/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 697519.1875\n",
      "Epoch 64/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 692843.8558\n",
      "Epoch 65/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 687993.1226\n",
      "Epoch 66/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 683018.8868\n",
      "Epoch 67/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 677895.1207\n",
      "Epoch 68/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 672572.8576\n",
      "Epoch 69/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 667116.6062\n",
      "Epoch 70/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 661493.6549\n",
      "Epoch 71/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 655727.3043\n",
      "Epoch 72/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 649785.7996\n",
      "Epoch 73/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 643658.0056\n",
      "Epoch 74/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 637365.2151\n",
      "Epoch 75/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 630905.3330\n",
      "Epoch 76/250\n",
      "670/670 [==============================] - 0s 48us/sample - loss: 624258.9916\n",
      "Epoch 77/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 617469.4450\n",
      "Epoch 78/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 610495.9455\n",
      "Epoch 79/250\n",
      "670/670 [==============================] - 0s 49us/sample - loss: 603352.1703\n",
      "Epoch 80/250\n",
      "670/670 [==============================] - 0s 49us/sample - loss: 596030.7993\n",
      "Epoch 81/250\n",
      "670/670 [==============================] - 0s 51us/sample - loss: 588542.4632\n",
      "Epoch 82/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 580859.4065\n",
      "Epoch 83/250\n",
      "670/670 [==============================] - 0s 49us/sample - loss: 572990.3629\n",
      "Epoch 84/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 564967.0845\n",
      "Epoch 85/250\n",
      "670/670 [==============================] - 0s 48us/sample - loss: 556769.9606\n",
      "Epoch 86/250\n",
      "670/670 [==============================] - 0s 49us/sample - loss: 548439.9477\n",
      "Epoch 87/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 539892.7847\n",
      "Epoch 88/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 531185.6871\n",
      "Epoch 89/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 522294.5325\n",
      "Epoch 90/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "670/670 [==============================] - 0s 45us/sample - loss: 513208.3162\n",
      "Epoch 91/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 504007.3252\n",
      "Epoch 92/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 494643.2957\n",
      "Epoch 93/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 485123.4325\n",
      "Epoch 94/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 475444.5193\n",
      "Epoch 95/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 465645.8094\n",
      "Epoch 96/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 455646.3583\n",
      "Epoch 97/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 445530.1311\n",
      "Epoch 98/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 435242.2110\n",
      "Epoch 99/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 424763.1027\n",
      "Epoch 100/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 414261.7255\n",
      "Epoch 101/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 403656.9188\n",
      "Epoch 102/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 392866.0349\n",
      "Epoch 103/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 381942.4045\n",
      "Epoch 104/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 370985.5558\n",
      "Epoch 105/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 359942.3154\n",
      "Epoch 106/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 348744.3722\n",
      "Epoch 107/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 337507.5516\n",
      "Epoch 108/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 326198.8317\n",
      "Epoch 109/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 314761.3296\n",
      "Epoch 110/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 303359.6646\n",
      "Epoch 111/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 291909.4955\n",
      "Epoch 112/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 280413.4076\n",
      "Epoch 113/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 268965.2194\n",
      "Epoch 114/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 257399.2185\n",
      "Epoch 115/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 245914.6162\n",
      "Epoch 116/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 234374.5533\n",
      "Epoch 117/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 223024.8431\n",
      "Epoch 118/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 211626.2125\n",
      "Epoch 119/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 200376.2067\n",
      "Epoch 120/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 189159.9913\n",
      "Epoch 121/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 178057.4438\n",
      "Epoch 122/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 167092.4275\n",
      "Epoch 123/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 156276.5184\n",
      "Epoch 124/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 145674.6355\n",
      "Epoch 125/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 135238.3282\n",
      "Epoch 126/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 124986.1134\n",
      "Epoch 127/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 114988.7749\n",
      "Epoch 128/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 105202.3198\n",
      "Epoch 129/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 95729.0172\n",
      "Epoch 130/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 86557.1341\n",
      "Epoch 131/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 77740.1852\n",
      "Epoch 132/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 69313.0344\n",
      "Epoch 133/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 61264.1838\n",
      "Epoch 134/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 53613.4472\n",
      "Epoch 135/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 46361.5102\n",
      "Epoch 136/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 39631.3525\n",
      "Epoch 137/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 33476.5491\n",
      "Epoch 138/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 27740.5828\n",
      "Epoch 139/250\n",
      "670/670 [==============================] - 0s 48us/sample - loss: 22657.9357\n",
      "Epoch 140/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 18168.0020\n",
      "Epoch 141/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 14320.9793\n",
      "Epoch 142/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 11116.2304\n",
      "Epoch 143/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 8590.6935\n",
      "Epoch 144/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 6790.5060\n",
      "Epoch 145/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 5598.0077\n",
      "Epoch 146/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 4983.7197\n",
      "Epoch 147/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 4780.3340\n",
      "Epoch 148/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 4685.2237\n",
      "Epoch 149/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 4622.9572\n",
      "Epoch 150/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 4566.7235\n",
      "Epoch 151/250\n",
      "670/670 [==============================] - 0s 41us/sample - loss: 4510.1676\n",
      "Epoch 152/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 4446.1679\n",
      "Epoch 153/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 4394.5234\n",
      "Epoch 154/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 4338.4721\n",
      "Epoch 155/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 4283.6892\n",
      "Epoch 156/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 4225.5519\n",
      "Epoch 157/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 4168.0398\n",
      "Epoch 158/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 4119.6026\n",
      "Epoch 159/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 4062.1710\n",
      "Epoch 160/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 4006.8744\n",
      "Epoch 161/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 3945.5612\n",
      "Epoch 162/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 3893.1826\n",
      "Epoch 163/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 3844.2296\n",
      "Epoch 164/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 3788.6542\n",
      "Epoch 165/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 3743.8045\n",
      "Epoch 166/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3692.6758\n",
      "Epoch 167/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3641.2972\n",
      "Epoch 168/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 3582.9219\n",
      "Epoch 169/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3524.5083\n",
      "Epoch 170/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 3474.3502\n",
      "Epoch 171/250\n",
      "670/670 [==============================] - 0s 47us/sample - loss: 3419.5269\n",
      "Epoch 172/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 3373.7783\n",
      "Epoch 173/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 3318.7095\n",
      "Epoch 174/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3272.9766\n",
      "Epoch 175/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3219.3947\n",
      "Epoch 176/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3162.8225\n",
      "Epoch 177/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3110.4429\n",
      "Epoch 178/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 3061.9454\n",
      "Epoch 179/250\n",
      "670/670 [==============================] - 0s 49us/sample - loss: 3009.5738\n",
      "Epoch 180/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "670/670 [==============================] - 0s 48us/sample - loss: 2969.4847\n",
      "Epoch 181/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 2917.0121\n",
      "Epoch 182/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 2866.2660\n",
      "Epoch 183/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 2819.5755\n",
      "Epoch 184/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2768.6268\n",
      "Epoch 185/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 2726.2738\n",
      "Epoch 186/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 2678.9498\n",
      "Epoch 187/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 2632.4840\n",
      "Epoch 188/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 2593.1119\n",
      "Epoch 189/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2543.5455\n",
      "Epoch 190/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 2494.6992\n",
      "Epoch 191/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 2446.0324\n",
      "Epoch 192/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2400.3819\n",
      "Epoch 193/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2354.9681\n",
      "Epoch 194/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2312.3834\n",
      "Epoch 195/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 2269.3642\n",
      "Epoch 196/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2222.5589\n",
      "Epoch 197/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 2185.7161\n",
      "Epoch 198/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2146.1762\n",
      "Epoch 199/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2102.0193\n",
      "Epoch 200/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 2058.0423\n",
      "Epoch 201/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 2018.5934\n",
      "Epoch 202/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 1970.5486\n",
      "Epoch 203/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 1935.4683\n",
      "Epoch 204/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1895.2977\n",
      "Epoch 205/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1853.8584\n",
      "Epoch 206/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1820.3311\n",
      "Epoch 207/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 1776.0040\n",
      "Epoch 208/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 1736.4406\n",
      "Epoch 209/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1703.7298\n",
      "Epoch 210/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1666.7358\n",
      "Epoch 211/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1632.6528\n",
      "Epoch 212/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 1602.3476\n",
      "Epoch 213/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1566.9085\n",
      "Epoch 214/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 1529.8778\n",
      "Epoch 215/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1492.9190\n",
      "Epoch 216/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1456.0330\n",
      "Epoch 217/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 1426.7091\n",
      "Epoch 218/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 1388.5013\n",
      "Epoch 219/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1356.4272\n",
      "Epoch 220/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 1328.0917\n",
      "Epoch 221/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1299.2200\n",
      "Epoch 222/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1269.0792\n",
      "Epoch 223/250\n",
      "670/670 [==============================] - 0s 46us/sample - loss: 1238.3376\n",
      "Epoch 224/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1207.1759\n",
      "Epoch 225/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1175.7668\n",
      "Epoch 226/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1145.5603\n",
      "Epoch 227/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1114.2222\n",
      "Epoch 228/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1081.8449\n",
      "Epoch 229/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 1048.7547\n",
      "Epoch 230/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 1012.3250\n",
      "Epoch 231/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 991.0450\n",
      "Epoch 232/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 960.0577\n",
      "Epoch 233/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 930.9779\n",
      "Epoch 234/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 903.0204\n",
      "Epoch 235/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 874.9695\n",
      "Epoch 236/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 849.5029\n",
      "Epoch 237/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 821.5590\n",
      "Epoch 238/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 796.6845\n",
      "Epoch 239/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 772.3822\n",
      "Epoch 240/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 749.6316\n",
      "Epoch 241/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 725.3966\n",
      "Epoch 242/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 701.2286\n",
      "Epoch 243/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 681.8695\n",
      "Epoch 244/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 657.6644\n",
      "Epoch 245/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 636.5654\n",
      "Epoch 246/250\n",
      "670/670 [==============================] - 0s 45us/sample - loss: 615.0685\n",
      "Epoch 247/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 593.7592\n",
      "Epoch 248/250\n",
      "670/670 [==============================] - 0s 42us/sample - loss: 572.3096\n",
      "Epoch 249/250\n",
      "670/670 [==============================] - 0s 43us/sample - loss: 550.9207\n",
      "Epoch 250/250\n",
      "670/670 [==============================] - 0s 44us/sample - loss: 528.9662\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7ffbb1073410>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(x_train,y_train,epochs=250)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss = model.history.history[\"loss\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7ffbc34c0b10>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sbn.lineplot(x=range(len(loss)),y=loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainLoss = model.evaluate(x_train,y_train, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "testLoss = model.evaluate(x_test,y_test,verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "516.4895447688316"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainLoss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "506.748486328125"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testLoss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "testTahminleri = model.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1056.5616 ],\n",
       "       [ 659.1396 ],\n",
       "       [ 879.8033 ],\n",
       "       [ 897.2417 ],\n",
       "       [ 902.1072 ],\n",
       "       [ 510.9165 ],\n",
       "       [ 928.5295 ],\n",
       "       [ 981.5555 ],\n",
       "       [ 935.8912 ],\n",
       "       [1003.3185 ],\n",
       "       [ 770.24536],\n",
       "       [ 915.6222 ],\n",
       "       [ 943.631  ],\n",
       "       [1036.117  ],\n",
       "       [1093.5881 ],\n",
       "       [ 710.34106],\n",
       "       [1100.7997 ],\n",
       "       [ 684.7312 ],\n",
       "       [1125.1798 ],\n",
       "       [ 889.00104],\n",
       "       [ 894.81604],\n",
       "       [ 739.6898 ],\n",
       "       [ 512.04803],\n",
       "       [ 650.5987 ],\n",
       "       [ 843.5318 ],\n",
       "       [1067.7294 ],\n",
       "       [ 738.5338 ],\n",
       "       [ 776.31476],\n",
       "       [ 878.6393 ],\n",
       "       [ 788.24274],\n",
       "       [ 509.20975],\n",
       "       [ 798.1173 ],\n",
       "       [ 751.0147 ],\n",
       "       [ 680.25684],\n",
       "       [ 889.781  ],\n",
       "       [ 852.129  ],\n",
       "       [1019.52167],\n",
       "       [1009.46173],\n",
       "       [ 852.4792 ],\n",
       "       [ 868.50354],\n",
       "       [ 775.8998 ],\n",
       "       [1142.0702 ],\n",
       "       [1073.9359 ],\n",
       "       [1065.1575 ],\n",
       "       [ 876.2076 ],\n",
       "       [ 635.56024],\n",
       "       [1117.6466 ],\n",
       "       [ 965.95917],\n",
       "       [1025.7764 ],\n",
       "       [ 863.8151 ],\n",
       "       [ 817.5763 ],\n",
       "       [ 866.1225 ],\n",
       "       [ 809.2826 ],\n",
       "       [ 997.02405],\n",
       "       [1020.51276],\n",
       "       [ 930.26733],\n",
       "       [ 943.30493],\n",
       "       [1056.469  ],\n",
       "       [ 887.23145],\n",
       "       [ 998.29486],\n",
       "       [ 802.9039 ],\n",
       "       [ 857.82007],\n",
       "       [ 750.2841 ],\n",
       "       [ 897.15173],\n",
       "       [ 962.81805],\n",
       "       [ 727.25806],\n",
       "       [ 808.62134],\n",
       "       [ 821.1229 ],\n",
       "       [ 863.8717 ],\n",
       "       [ 837.4897 ],\n",
       "       [ 816.45984],\n",
       "       [ 935.71423],\n",
       "       [1040.7543 ],\n",
       "       [ 941.2947 ],\n",
       "       [ 978.8505 ],\n",
       "       [ 697.2972 ],\n",
       "       [ 672.57117],\n",
       "       [ 725.458  ],\n",
       "       [ 883.07935],\n",
       "       [ 909.0529 ],\n",
       "       [ 973.28467],\n",
       "       [ 979.1114 ],\n",
       "       [ 650.3895 ],\n",
       "       [ 868.9354 ],\n",
       "       [ 659.914  ],\n",
       "       [ 951.705  ],\n",
       "       [ 837.5485 ],\n",
       "       [ 897.8516 ],\n",
       "       [ 880.1169 ],\n",
       "       [ 666.31866],\n",
       "       [ 809.7215 ],\n",
       "       [ 734.9023 ],\n",
       "       [1158.6389 ],\n",
       "       [ 895.2522 ],\n",
       "       [ 839.77893],\n",
       "       [1076.6494 ],\n",
       "       [ 860.39795],\n",
       "       [ 958.9855 ],\n",
       "       [ 959.4828 ],\n",
       "       [ 497.34137],\n",
       "       [ 977.1024 ],\n",
       "       [1044.0244 ],\n",
       "       [ 884.8096 ],\n",
       "       [1037.4104 ],\n",
       "       [ 605.85455],\n",
       "       [ 695.20355],\n",
       "       [ 863.82245],\n",
       "       [ 735.662  ],\n",
       "       [ 918.23553],\n",
       "       [ 924.9555 ],\n",
       "       [ 743.69867],\n",
       "       [ 886.02716],\n",
       "       [ 971.87213],\n",
       "       [1041.9927 ],\n",
       "       [ 720.40643],\n",
       "       [ 908.32574],\n",
       "       [ 863.1024 ],\n",
       "       [ 872.82855],\n",
       "       [1017.2882 ],\n",
       "       [ 802.94617],\n",
       "       [ 736.93365],\n",
       "       [ 867.45605],\n",
       "       [1102.93   ],\n",
       "       [ 815.3058 ],\n",
       "       [ 833.4117 ],\n",
       "       [ 899.2485 ],\n",
       "       [ 790.9598 ],\n",
       "       [ 929.7998 ],\n",
       "       [ 955.9121 ],\n",
       "       [ 921.13025],\n",
       "       [ 944.34015],\n",
       "       [1074.9266 ],\n",
       "       [1015.5833 ],\n",
       "       [ 814.3698 ],\n",
       "       [ 775.9657 ],\n",
       "       [ 789.22906],\n",
       "       [ 758.693  ],\n",
       "       [ 761.78033],\n",
       "       [ 705.9961 ],\n",
       "       [1109.1686 ],\n",
       "       [ 682.40204],\n",
       "       [ 972.4132 ],\n",
       "       [ 731.94696],\n",
       "       [ 895.6511 ],\n",
       "       [ 976.4963 ],\n",
       "       [ 642.7431 ],\n",
       "       [ 763.5126 ],\n",
       "       [ 802.3769 ],\n",
       "       [ 873.75336],\n",
       "       [ 767.946  ],\n",
       "       [ 600.7324 ],\n",
       "       [ 668.7455 ],\n",
       "       [ 754.5198 ],\n",
       "       [1038.82   ],\n",
       "       [1085.0259 ],\n",
       "       [ 901.99274],\n",
       "       [ 739.1938 ],\n",
       "       [ 803.5387 ],\n",
       "       [ 884.0818 ],\n",
       "       [ 950.6797 ],\n",
       "       [ 638.7335 ],\n",
       "       [ 858.8715 ],\n",
       "       [ 885.37024],\n",
       "       [ 669.0601 ],\n",
       "       [ 915.6676 ],\n",
       "       [ 987.3288 ],\n",
       "       [1150.0565 ],\n",
       "       [ 815.8171 ],\n",
       "       [ 761.1082 ],\n",
       "       [ 955.397  ],\n",
       "       [ 739.13086],\n",
       "       [1002.26855],\n",
       "       [1186.8816 ],\n",
       "       [ 902.268  ],\n",
       "       [ 884.73883],\n",
       "       [ 682.6966 ],\n",
       "       [ 648.6464 ],\n",
       "       [ 762.5622 ],\n",
       "       [ 956.4679 ],\n",
       "       [ 727.104  ],\n",
       "       [ 782.75757],\n",
       "       [ 827.8094 ],\n",
       "       [ 719.19775],\n",
       "       [ 920.0472 ],\n",
       "       [ 877.41815],\n",
       "       [ 649.69525],\n",
       "       [1042.5173 ],\n",
       "       [ 941.52637],\n",
       "       [1118.592  ],\n",
       "       [ 802.10364],\n",
       "       [ 882.12036],\n",
       "       [ 688.22864],\n",
       "       [ 697.9797 ],\n",
       "       [ 755.6931 ],\n",
       "       [ 712.301  ],\n",
       "       [1060.1807 ],\n",
       "       [ 800.0061 ],\n",
       "       [ 862.74384],\n",
       "       [1115.7778 ],\n",
       "       [ 626.33374],\n",
       "       [ 950.4409 ],\n",
       "       [ 693.03625],\n",
       "       [ 612.0618 ],\n",
       "       [ 767.9926 ],\n",
       "       [ 634.6648 ],\n",
       "       [ 991.8925 ],\n",
       "       [1018.63873],\n",
       "       [ 667.37933],\n",
       "       [ 934.41595],\n",
       "       [ 839.2923 ],\n",
       "       [1107.7338 ],\n",
       "       [ 917.4961 ],\n",
       "       [ 890.60284],\n",
       "       [ 968.115  ],\n",
       "       [1019.18085],\n",
       "       [ 880.5014 ],\n",
       "       [ 858.2863 ],\n",
       "       [ 846.88184],\n",
       "       [ 684.24695],\n",
       "       [1293.6765 ],\n",
       "       [ 922.1528 ],\n",
       "       [ 580.1184 ],\n",
       "       [ 811.5717 ],\n",
       "       [ 981.165  ],\n",
       "       [ 862.774  ],\n",
       "       [ 806.30396],\n",
       "       [ 706.59955],\n",
       "       [ 805.9211 ],\n",
       "       [ 963.729  ],\n",
       "       [ 917.1161 ],\n",
       "       [ 987.60284],\n",
       "       [ 939.90173],\n",
       "       [ 905.32336],\n",
       "       [ 763.0122 ],\n",
       "       [1041.522  ],\n",
       "       [1111.8889 ],\n",
       "       [ 998.3908 ],\n",
       "       [1001.7637 ],\n",
       "       [ 984.22   ],\n",
       "       [ 858.842  ],\n",
       "       [ 997.6342 ],\n",
       "       [ 662.6374 ],\n",
       "       [ 896.58435],\n",
       "       [1046.9061 ],\n",
       "       [ 797.484  ],\n",
       "       [ 610.61035],\n",
       "       [ 646.13904],\n",
       "       [ 979.5332 ],\n",
       "       [ 882.21063],\n",
       "       [1050.1873 ],\n",
       "       [ 855.2921 ],\n",
       "       [ 764.2433 ],\n",
       "       [ 589.93854],\n",
       "       [ 878.46136],\n",
       "       [ 796.9475 ],\n",
       "       [ 858.31854],\n",
       "       [ 891.268  ],\n",
       "       [ 911.0632 ],\n",
       "       [ 653.37354],\n",
       "       [1007.0415 ],\n",
       "       [ 664.6481 ],\n",
       "       [ 888.1554 ],\n",
       "       [ 812.1065 ],\n",
       "       [ 915.1903 ],\n",
       "       [ 960.65875],\n",
       "       [ 804.40063],\n",
       "       [ 881.8307 ],\n",
       "       [1127.2247 ],\n",
       "       [ 997.0945 ],\n",
       "       [ 856.3711 ],\n",
       "       [1061.1833 ],\n",
       "       [ 985.8788 ],\n",
       "       [ 626.1895 ],\n",
       "       [ 592.5453 ],\n",
       "       [ 965.17053],\n",
       "       [1030.0125 ],\n",
       "       [ 942.6145 ],\n",
       "       [ 904.0824 ],\n",
       "       [ 464.71872],\n",
       "       [1002.58685],\n",
       "       [ 914.34705],\n",
       "       [ 831.32556],\n",
       "       [ 931.80005],\n",
       "       [ 924.9271 ],\n",
       "       [ 976.9756 ],\n",
       "       [ 965.7061 ],\n",
       "       [ 956.02   ],\n",
       "       [ 652.86743],\n",
       "       [ 963.6917 ],\n",
       "       [ 969.144  ],\n",
       "       [ 884.3524 ],\n",
       "       [ 861.70776],\n",
       "       [ 755.4345 ],\n",
       "       [ 944.37195],\n",
       "       [ 892.2811 ],\n",
       "       [ 819.263  ],\n",
       "       [ 961.9453 ],\n",
       "       [ 844.5227 ],\n",
       "       [1001.21454],\n",
       "       [ 643.3577 ],\n",
       "       [ 615.07367],\n",
       "       [ 744.8042 ],\n",
       "       [ 818.3705 ],\n",
       "       [ 889.0704 ],\n",
       "       [1182.7671 ],\n",
       "       [ 647.0452 ],\n",
       "       [1023.28076],\n",
       "       [ 695.6664 ],\n",
       "       [1099.9108 ],\n",
       "       [ 813.7502 ],\n",
       "       [1176.197  ],\n",
       "       [ 803.8752 ],\n",
       "       [ 676.588  ],\n",
       "       [ 897.6571 ],\n",
       "       [ 921.9697 ],\n",
       "       [ 851.9404 ],\n",
       "       [ 697.5155 ],\n",
       "       [ 978.6907 ],\n",
       "       [1068.2128 ],\n",
       "       [ 792.7142 ],\n",
       "       [ 897.32227],\n",
       "       [1038.7246 ],\n",
       "       [ 970.32355],\n",
       "       [ 736.1374 ],\n",
       "       [ 645.62714],\n",
       "       [1000.0597 ],\n",
       "       [ 806.4288 ],\n",
       "       [ 783.23804],\n",
       "       [1134.7095 ],\n",
       "       [ 817.3614 ]], dtype=float32)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testTahminleri"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "tahminDf = pd.DataFrame(y_test,columns=[\"Gerçek Y\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gerçek Y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1081.652164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>622.675990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>889.356810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>902.826733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>897.662404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>1028.438035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>789.934950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>327</th>\n",
       "      <td>758.490486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>1172.871659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>820.947936</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>330 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Gerçek Y\n",
       "0    1081.652164\n",
       "1     622.675990\n",
       "2     889.356810\n",
       "3     902.826733\n",
       "4     897.662404\n",
       "..           ...\n",
       "325  1028.438035\n",
       "326   789.934950\n",
       "327   758.490486\n",
       "328  1172.871659\n",
       "329   820.947936\n",
       "\n",
       "[330 rows x 1 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tahminDf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "testTahminleri = pd.Series(testTahminleri.reshape(330,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      1056.561646\n",
       "1       659.139587\n",
       "2       879.803284\n",
       "3       897.241699\n",
       "4       902.107178\n",
       "          ...     \n",
       "325    1000.059692\n",
       "326     806.428772\n",
       "327     783.238037\n",
       "328    1134.709473\n",
       "329     817.361389\n",
       "Length: 330, dtype: float32"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testTahminleri"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "tahminDf = pd.concat([tahminDf,testTahminleri],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gerçek Y</th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1081.652164</td>\n",
       "      <td>1056.561646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>622.675990</td>\n",
       "      <td>659.139587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>889.356810</td>\n",
       "      <td>879.803284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>902.826733</td>\n",
       "      <td>897.241699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>897.662404</td>\n",
       "      <td>902.107178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>1028.438035</td>\n",
       "      <td>1000.059692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>789.934950</td>\n",
       "      <td>806.428772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>327</th>\n",
       "      <td>758.490486</td>\n",
       "      <td>783.238037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>1172.871659</td>\n",
       "      <td>1134.709473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>820.947936</td>\n",
       "      <td>817.361389</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>330 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Gerçek Y            0\n",
       "0    1081.652164  1056.561646\n",
       "1     622.675990   659.139587\n",
       "2     889.356810   879.803284\n",
       "3     902.826733   897.241699\n",
       "4     897.662404   902.107178\n",
       "..           ...          ...\n",
       "325  1028.438035  1000.059692\n",
       "326   789.934950   806.428772\n",
       "327   758.490486   783.238037\n",
       "328  1172.871659  1134.709473\n",
       "329   820.947936   817.361389\n",
       "\n",
       "[330 rows x 2 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tahminDf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "tahminDf.columns = [\"Gerçek Y\", \"Tahmin Y\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gerçek Y</th>\n",
       "      <th>Tahmin Y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1081.652164</td>\n",
       "      <td>1056.561646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>622.675990</td>\n",
       "      <td>659.139587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>889.356810</td>\n",
       "      <td>879.803284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>902.826733</td>\n",
       "      <td>897.241699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>897.662404</td>\n",
       "      <td>902.107178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>1028.438035</td>\n",
       "      <td>1000.059692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>789.934950</td>\n",
       "      <td>806.428772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>327</th>\n",
       "      <td>758.490486</td>\n",
       "      <td>783.238037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>1172.871659</td>\n",
       "      <td>1134.709473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>820.947936</td>\n",
       "      <td>817.361389</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>330 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Gerçek Y     Tahmin Y\n",
       "0    1081.652164  1056.561646\n",
       "1     622.675990   659.139587\n",
       "2     889.356810   879.803284\n",
       "3     902.826733   897.241699\n",
       "4     897.662404   902.107178\n",
       "..           ...          ...\n",
       "325  1028.438035  1000.059692\n",
       "326   789.934950   806.428772\n",
       "327   758.490486   783.238037\n",
       "328  1172.871659  1134.709473\n",
       "329   820.947936   817.361389\n",
       "\n",
       "[330 rows x 2 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tahminDf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7ffbb12954d0>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sbn.scatterplot(x = \"Gerçek Y\", y = \"Tahmin Y\", data = tahminDf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_absolute_error, mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18.037772147916666"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_absolute_error(tahminDf[\"Gerçek Y\"],tahminDf[\"Tahmin Y\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "506.7484460343649"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(tahminDf[\"Gerçek Y\"],tahminDf[\"Tahmin Y\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Fiyat</th>\n",
       "      <th>BisikletOzellik1</th>\n",
       "      <th>BisikletOzellik2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>872.677801</td>\n",
       "      <td>1750.024800</td>\n",
       "      <td>1749.964733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>164.124504</td>\n",
       "      <td>1.704531</td>\n",
       "      <td>1.659578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>390.856887</td>\n",
       "      <td>1744.852108</td>\n",
       "      <td>1744.742389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>757.795031</td>\n",
       "      <td>1748.831119</td>\n",
       "      <td>1748.803186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>879.168705</td>\n",
       "      <td>1750.017350</td>\n",
       "      <td>1750.003926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>988.612778</td>\n",
       "      <td>1751.115766</td>\n",
       "      <td>1751.129414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1355.213745</td>\n",
       "      <td>1755.613884</td>\n",
       "      <td>1754.666038</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Fiyat  BisikletOzellik1  BisikletOzellik2\n",
       "count  1000.000000       1000.000000       1000.000000\n",
       "mean    872.677801       1750.024800       1749.964733\n",
       "std     164.124504          1.704531          1.659578\n",
       "min     390.856887       1744.852108       1744.742389\n",
       "25%     757.795031       1748.831119       1748.803186\n",
       "50%     879.168705       1750.017350       1750.003926\n",
       "75%     988.612778       1751.115766       1751.129414\n",
       "max    1355.213745       1755.613884       1754.666038"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataFrame.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "yeniBisikletOzellikleri = [[1751,1750]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "yeniBisikletOzellikleri = scaler.transform(yeniBisikletOzellikleri)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[913.2747]], dtype=float32)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(yeniBisikletOzellikleri)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import load_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save(\"bisiklet_modeli.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Sequential models without an `input_shape` passed to the first layer cannot reload their optimizer state. As a result, your model isstarting with a freshly initialized optimizer.\n"
     ]
    }
   ],
   "source": [
    "sonradanCagirilanModel= load_model(\"bisiklet_modeli.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[913.2747]], dtype=float32)"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sonradanCagirilanModel.predict(yeniBisikletOzellikleri)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}