{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "torch.set_printoptions(edgeitems=2, threshold=50, linewidth=75)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[1.0000e+00, 1.0000e+00,  ..., 1.3000e+01, 1.6000e+01],\n",
       "        [2.0000e+00, 1.0000e+00,  ..., 3.2000e+01, 4.0000e+01],\n",
       "        ...,\n",
       "        [1.7378e+04, 3.1000e+01,  ..., 4.8000e+01, 6.1000e+01],\n",
       "        [1.7379e+04, 3.1000e+01,  ..., 3.7000e+01, 4.9000e+01]])"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bikes_numpy = np.loadtxt(\n",
    "    \"../data/p1ch4/bike-sharing-dataset/hour-fixed.csv\", \n",
    "    dtype=np.float32, \n",
    "    delimiter=\",\", \n",
    "    skiprows=1, \n",
    "    converters={1: lambda x: float(x[8:10])}) # <1>\n",
    "bikes = torch.from_numpy(bikes_numpy)\n",
    "bikes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([17520, 17]), (17, 1))"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bikes.shape, bikes.stride()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([730, 24, 17]), (408, 17, 1))"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "daily_bikes = bikes.view(-1, 24, bikes.shape[1])\n",
    "daily_bikes.shape, daily_bikes.stride()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([730, 17, 24]), (408, 1, 17))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "daily_bikes = daily_bikes.transpose(1, 2)\n",
    "daily_bikes.shape, daily_bikes.stride()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 3, 3, 2, 2,\n",
       "        2, 2])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first_day = bikes[:24].long()\n",
    "weather_onehot = torch.zeros(first_day.shape[0], 4)\n",
    "first_day[:,9]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[1., 0., 0., 0.],\n",
       "        [1., 0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 1., 0., 0.],\n",
       "        [0., 1., 0., 0.]])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weather_onehot.scatter_(\n",
    "    dim=1, \n",
    "    index=first_day[:,9].unsqueeze(1).long() - 1, # <1>\n",
    "    value=1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 1.0000,  1.0000,  1.0000,  0.0000,  1.0000,  0.0000,  0.0000,\n",
       "          6.0000,  0.0000,  1.0000,  0.2400,  0.2879,  0.8100,  0.0000,\n",
       "          3.0000, 13.0000, 16.0000,  1.0000,  0.0000,  0.0000,  0.0000]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.cat((bikes[:24], weather_onehot), 1)[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([730, 4, 24])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "daily_weather_onehot = torch.zeros(daily_bikes.shape[0], 4,\n",
    "                                   daily_bikes.shape[2])\n",
    "daily_weather_onehot.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([730, 4, 24])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "daily_weather_onehot.scatter_(\n",
    "    1, daily_bikes[:,9,:].long().unsqueeze(1) - 1, 1.0)\n",
    "daily_weather_onehot.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "daily_bikes = torch.cat((daily_bikes, daily_weather_onehot), dim=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "daily_bikes[:, 9, :] = (daily_bikes[:, 9, :] - 1.0) / 3.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "temp = daily_bikes[:, 10, :]\n",
    "temp_min = torch.min(temp)\n",
    "temp_max = torch.max(temp)\n",
    "daily_bikes[:, 10, :] = ((daily_bikes[:, 10, :] - temp_min)\n",
    "                         / (temp_max - temp_min))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "temp = daily_bikes[:, 10, :]\n",
    "daily_bikes[:, 10, :] = ((daily_bikes[:, 10, :] - torch.mean(temp))\n",
    "                         / torch.std(temp))"
   ]
  }
 ],
 "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": 2
}