PyTorch5_Time_Series_Analysis.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "PyTorch5_Time_Series_Analysis.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyMpkVYGxmfqzBn8TRcBhCGi",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/utkarsh0702/Pytorch/blob/master/PyTorch5_Time_Series_Analysis.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0RjbvP2_YxhZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import torch\n",
        "from torch import nn\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EiQq58NVZ0WV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        },
        "outputId": "1c1b785e-935e-4a1f-c3b1-bb26ba83f9be"
      },
      "source": [
        "plt.figure(figsize=(8, 6))\n",
        "seq_length= 20\n",
        "time_step= np.linspace(0, np.pi, seq_length+1)\n",
        "data= np.sin(time_step)\n",
        "data.resize((seq_length+1,1))\n",
        "\n",
        "x= data[:-1]\n",
        "y= data[1:]\n",
        "\n",
        "plt.plot(time_step[1:], x, \"r.\", label=\"Input(x)\")\n",
        "plt.plot(time_step[1:], y, \"b.\", label=\"Target(y)\")\n",
        "plt.show()"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "55jrr27BbQNV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class RNN(nn.Module):\n",
        "  def __init__(self, input_size, output_size, hidden_dim, n_layers):\n",
        "    super(RNN, self).__init__()\n",
        "    self.hidden_dim= hidden_dim\n",
        "\n",
        "    self.rnn= nn.RNN(input_size, hidden_dim, n_layers, batch_first=True)\n",
        "    self.fc= nn.Linear(hidden_dim, output_size)\n",
        "  \n",
        "  def forward(self, x, hidden):\n",
        "    batch_size= x.size(0)\n",
        "    r_out, hidden= self.rnn(x, hidden)\n",
        "    r_out= r_out.view(-1, self.hidden_dim)\n",
        "\n",
        "    output= self.fc(r_out)\n",
        "    return output, hidden"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3Q-IU6LfdOuD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "b2d76694-2f3e-47cb-e2cc-f7a8d5c88885"
      },
      "source": [
        "#Test the RNN for Shape and Size\n",
        "test_rnn= RNN(input_size=1, output_size=1, hidden_dim=10, n_layers=2)\n",
        "\n",
        "#Evenly spaced test data points\n",
        "time_step= np.linspace(0, np.pi, seq_length)\n",
        "data= np.sin(time_step)\n",
        "data.resize((seq_length,1))\n",
        "\n",
        "test_input= torch.Tensor(data).unsqueeze(0)\n",
        "print('Input Size:', test_input.size())\n",
        "\n",
        "test_out, test_h= test_rnn(test_input, None)\n",
        "print('Output Size:', test_out.size())\n",
        "print('Hidden State Size:', test_h.size())"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Input Size: torch.Size([1, 20, 1])\n",
            "Output Size: torch.Size([20, 1])\n",
            "Hidden State Size: torch.Size([2, 1, 10])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QVYOEM9zeggE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "e61ac336-1023-4ad8-aff4-b63d50587265"
      },
      "source": [
        "#Instansiate an RNN\n",
        "rnn= RNN(1,1,32,1)\n",
        "print(rnn)\n",
        "loss_function= nn.MSELoss()\n",
        "optimizer= torch.optim.Adam(rnn.parameters(), lr=0.01)"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "RNN(\n",
            "  (rnn): RNN(1, 32, batch_first=True)\n",
            "  (fc): Linear(in_features=32, out_features=1, bias=True)\n",
            ")\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nm0Cy0cjfQNF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train(rnn, n_steps, print_every):\n",
        "  hidden= None\n",
        "  for i, step in enumerate(range(n_steps)):\n",
        "    time_steps= np.linspace(step*np.pi, (step+1)*np.pi, seq_length+1)\n",
        "    data= np.sin(time_steps)\n",
        "    data.resize((seq_length+1,1))\n",
        "\n",
        "    x= data[:-1]\n",
        "    y= data[1:]\n",
        "    x_tensor= torch.Tensor(x).unsqueeze(0)\n",
        "    y_tensor= torch.Tensor(y)\n",
        "\n",
        "    prediction, hidden= rnn(x_tensor, hidden)\n",
        "    hidden= hidden.data\n",
        "    loss= loss_function(prediction, y_tensor)\n",
        "    optimizer.zero_grad()\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    if i% print_every ==0:\n",
        "      print('Loss:', loss.item())\n",
        "      plt.figure()\n",
        "      plt.plot(time_steps[1:],x,'r.')\n",
        "      plt.plot(time_steps[1:],prediction.data.numpy().flatten(),'b.')\n",
        "      plt.plot()\n",
        "  \n",
        "  return rnn"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wMP19003iXPU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "dd942db8-ec8a-4765-b463-629b31059d1c"
      },
      "source": [
        "n_steps=75\n",
        "print_every=15\n",
        "trained_rnn= train(rnn, n_steps, print_every)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Loss: 0.33332115411758423\n",
            "Loss: 0.015161627903580666\n",
            "Loss: 0.0007780947489663959\n",
            "Loss: 0.00039236946031451225\n",
            "Loss: 0.0005042075063101947\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD4CAYAAAD8Zh1EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAASGklEQVR4nO3df4wc913G8efpOW4iJSRSfIjIdntGuBImFJqsrJwioZPcICd/2H80IKcqbapQSyBDUSskB2haHMlWqShSqdVgkag/BHFDQNVRHFmVm1MkfAlet0mIbVwdhuIzkXwNqQsK7WHz4Y+ZC5vr7u3c7ezOznffL+m0P2Zu5zM7O89+9zu/HBECANTf26ouAABQDgIdABJBoANAIgh0AEgEgQ4AiVhX1YQ3bNgQExMTVU0eAGrp9OnT34uI8XbDKgv0iYkJNZvNqiYPALVk+7udhtHlAgCJINABIBEEOgAkgkAHgEQQ6ACQiK6BbvsJ25dtv9JhuG1/zvac7Zdt31F+mQCAboq00L8oaecKw++VtDX/2yvpC72XBQyp2Vnp0KHsFhgyXfdDj4jnbE+sMMpuSV+O7Dy8z9u+xfZtEfFqSTUCw2F2VtqxQ1pclNavl06ckCYnq64KeFMZfegbJV1seTyfP/djbO+13bTdXFhYKGHSwADNzGRhfu1adjszU3VFwFsMdKNoRByJiEZENMbH2x65CvRXL10mU1NZy3xsLLudmhp8DcAKyjj0/5KkzS2PN+XPAcOl1y6Tycnsf2ZmsjBfS3cL3TboozJa6NOSPpjv7XKXpCv0n2MoldFlMjkpPfzw2kOYbhv0UdcWuu0nJU1J2mB7XtInJV0nSRHxmKRjku6TNCfpDUkf7lexQE+WukyWWsdr7TKpew1Ilqu6SHSj0QjOtoiBm53trcsklRpQW7ZPR0Sj7TACHQDqY6VA59B/1At7iPAeoKPKLnABrBp7iPAeYEW00FEf7CHCe4AVEeioj7IO7Kkz3gOsgC4X1EcZB/bUHe8BVsBeLgBQI+zlAgAjgEAHgEQQ6ACQCAIdABJBoANAIgh0DBaHrVePZZAs9kPH4HDYevVYBkmjhY7B4bD16rEMkkagY3A4bL16LIOk0eWCweGw9eqxDJLGof8AUCMc+g8AI4BAB4BEEOgAkAgCHQASQaADQCIIdABIBIEOAIkg0AEgEQQ6ACSCQAeARBQKdNs7bZ+3PWd7f5vh77D9rO1v237Z9n3ll4qhwLm0wWdgaHU9OZftMUmHJd0jaV7SKdvTEXG2ZbQ/kPRURHzB9jZJxyRN9KFeVIlzaYPPwFAr0kLfLmkuIi5ExKKko5J2LxsnJP1Efv9mSf9eXokYGpxLG3wGhlqRQN8o6WLL4/n8uVafkvQB2/PKWue/1e6FbO+13bTdXFhYWEO5qBTn0gafgaFW1vnQH5D0xYj4Y9uTkr5i+/aI+N/WkSLiiKQjUnb63JKmjUHhXNrgMzDUigT6JUmbWx5vyp9r9ZCknZIUEbO2r5e0QdLlMorEEJmcZCUedXwGhlaRLpdTkrba3mJ7vaQ9kqaXjfNvknZIku2flXS9JPpUAGCAugZ6RFyVtE/ScUnnlO3Ncsb2Adu78tE+Lukjtl+S9KSkB6OqSyEBwIgq1IceEceUbexsfe6RlvtnJd1dbmkAgNXgSFEASASBDgCJINABIBEEOgAkgkAHgEQQ6ACQCAIdABJBoANAIgh0AEgEgT5quNoMqsZnsG/KOn0u6oCrzaBqfAb7ihb6KOFqM6gan8G+ItBHCVebQdX4DPYVXS6jhKvNoGp8BvvKVZ22vNFoRLPZrGTaAFBXtk9HRKPdMLpcACARBDoAJIJAB4BEEOgAkAgCHQASQaADQCIIdABIBIEOAIkg0AEgEQQ6ACSCQAeARBDoAJAIAh0AElEo0G3vtH3e9pzt/R3G+VXbZ22fsf2X5ZYJAOim6/nQbY9JOizpHknzkk7Zno6Isy3jbJX0sKS7I+J12z/Zr4IBAO0VaaFvlzQXERciYlHSUUm7l43zEUmHI+J1SYqIy+WWCQDopkigb5R0seXxfP5cq3dJepftv7f9vO2d7V7I9l7bTdvNhYWFtVU86rhiOkYd60BHZV2Cbp2krZKmJG2S9Jztn4+I77eOFBFHJB2RsisWlTTt0cEV0zHqWAdWVKSFfknS5pbHm/LnWs1Lmo6I/4mIf5H0HWUBjzJxxXSMOtaBFRUJ9FOSttreYnu9pD2SppeN8zVlrXPZ3qCsC+ZCiXVC4orpAOvAirp2uUTEVdv7JB2XNCbpiYg4Y/uApGZETOfDftn2WUnXJP1uRLzWz8JHEldMx6hjHViRI6rpym40GtFsNiuZNgDUle3TEdFoN4wjRQEgEQQ6ACSCQAeARBDoAJAIAh0AEkGgA0AiCHQASASBDgCJINABIBEEOgAkgkAHgEQQ6ACQCAIdABJBoANAIgh0AEgEgQ4AiSDQASARBDoAJIJAB4BEEOiDNjsrHTqU3QIYvITXwXVVFzBSZmelHTukxUVp/frs6uVctRwYnMTXQVrogzQzk32Qrl3Lbmdmqq4IGC2Jr4ME+iBNTWWtgrGx7HZqquqKgNGS+DpIl8sgTU5mP/FmZrIPUkI/9YBaSHwddERUMuFGoxHNZrOSaQNAXdk+HRGNdsPocgGARBDoAJAIAh0AElEo0G3vtH3e9pzt/SuM9z7bYbtt/w4AoH+6BrrtMUmHJd0raZukB2xvazPeTZI+KumFsosEAHRXpIW+XdJcRFyIiEVJRyXtbjPeo5I+LemHJdYHACioSKBvlHSx5fF8/tybbN8haXNE/N1KL2R7r+2m7ebCwsKqiwUAdNbzRlHbb5P0WUkf7zZuRByJiEZENMbHx3udNACgRZFAvyRpc8vjTflzS26SdLukGdv/KukuSdNsGAWAwSoS6KckbbW9xfZ6SXskTS8NjIgrEbEhIiYiYkLS85J2RQSHgQLAAHUN9Ii4KmmfpOOSzkl6KiLO2D5ge1e/CwQAFFPo5FwRcUzSsWXPPdJh3KneywIArBZHigJAIgh0AEgEgQ4AiSDQASARBDoAJIJAB4BEEOgAkAgCHQASQaADQCII9NWanZUOHcpuAYyeIc6AQof+Izc7K+3YIS0uSuvXSydOSJOTVVcFYFCGPANooa/GzEy2IK9dy25nZqquCMAgDXkGEOirMTWVfSuPjWW3U1NVVwRgkIY8A+hyWY3Jyewn1sxMtiCH6KcWgAEY8gxwRFQy4UajEc0m18AAgNWwfToi2l4Rji4XAEgEgQ4AiSDQASARBDoAJIJAB4BEEOgAkAgCHQASQaADQCIIdABIBIEOAIkg0AEgEQQ6ACSCQAeARBQKdNs7bZ+3PWd7f5vhH7N91vbLtk/Yfmf5pQIAVtI10G2PSTos6V5J2yQ9YHvbstG+LakREe+W9LSkPyq7UADAyoq00LdLmouICxGxKOmopN2tI0TEsxHxRv7weUmbyi0TANBNkUDfKOliy+P5/LlOHpL0TLsBtvfabtpuLiwsFK8SANBVqRtFbX9AUkPSZ9oNj4gjEdGIiMb4+HiZkwaAkVck0C9J2tzyeFP+3FvYfq+k35e0KyJ+VE55wFvNzkqHDmW3dfz/sl4DaKfIRaJPSdpqe4uyIN8j6f2tI9h+j6Q/k7QzIi6XXiWgLAB37JAWF7MLrp84sbpr9Fb9/2W+xpBeoxgV69pCj4irkvZJOi7pnKSnIuKM7QO2d+WjfUbSjZL+yvaLtqf7VjFqrZfW6cxMFoTXrmW3MzP1+v8yXmPpC+ETn8huaeWjVZEWuiLimKRjy557pOX+e0uuCwnqtXU6NZX939L/T02tbvpV/38Zr9HuC4EWPpYUCvSk8GnuSS9vX69hNDmZfQmsdfpV/38Zr9HrF0IZXT7oUT8zKCIq+bvzzjtj4E6ejLjhhoixsez25MnB11Bjvb59vP3lOHky4uDBtb1/Bw9m77+U3R48ONjpj7wSVgJJzeiQq6PVQi/j92rN1bmFjczk5NrfO1r4FetzBo1WoJfRCVpjVfdhS72FEXrX65cqbaIe9TmDRivQR7yJSAsbUrUtfGnEN2P1eSVy1iUzeI1GI5rNZiXTHlX8XEYZeglkPoO9s306Ihrtho1WCz0BvaxMtLBRhl5a+HTZ9BeBXiNltG7ow0aVRnwzVt9xxaIaKeNIRaBKS78SH3107d0tnAunM1roNULrBino5VciffArI9BrhD5wjDr64FdGoA9Yr7ts0QeOUcZukysj0AeIn4tAb3r9lZr6OkigDxA/F4HesdtkZ+zlMkBLPxfHxtioCVQh9XWQFvoAsVETqFbq6yCH/gPAKlS9UZVD/0tU9cIEUJ1h36hKoK/CsC9MAP017BtV2Si6Chx6D4y2Yd+oSgt9FTj0HhhtZWxU7We3LYG+CqlvIQfQ3TCfi4ZAXyUOvQewVv3ug6cPHQAGpN998LTQAWBA+t1tS6ADwAD1s9uWLhcASASBDgCJKBTotnfaPm97zvb+NsPfbvur+fAXbE+UXSgAYGVdA932mKTDku6VtE3SA7a3LRvtIUmvR8TPSPoTSZ8uu9A3cYVYAGiryEbR7ZLmIuKCJNk+Kmm3pLMt4+yW9Kn8/tOSPm/bUfapHDmZCgB0VKTLZaOkiy2P5/Pn2o4TEVclXZF06/IXsr3XdtN2c2FhYfXVcjIVAOhooBtFI+JIRDQiojE+Pr76Fxj2M+MAQIWKdLlckrS55fGm/Ll248zbXifpZkmvlVJhqxL2yud85gBSVSTQT0naanuLsuDeI+n9y8aZlvQhSbOS7pf0zdL7z5f0sFc+XfAAUta1yyXvE98n6bikc5Keiogztg/Y3pWP9rikW23PSfqYpB/btXEY0AUPIGWFDv2PiGOSji177pGW+z+U9CvlllY+zmcOIGUjdS4XzmcOIGUjFegS5zMHkC7O5QIAiSDQASARBDoAJKJ2gc65uQCgvVptFOXAIADorFYtdA4MAoDOahXonJsLADqrVZcLBwYBQGe1CnSJA4MAoJNadbkAADoj0AEgEQQ6ACSCQAeARBDoAJAIAh0AEuF+Xfqz64TtBUnfrWTi5dog6XtVF1ES5mU4MS/Dqap5eWdEjLcbUFmgp8J2MyIaVddRBuZlODEvw2kY54UuFwBIBIEOAIkg0Ht3pOoCSsS8DCfmZTgN3bzQhw4AiaCFDgCJINABIBEEekG2d9o+b3vO9v42wx+0vWD7xfzv16uosxvbT9i+bPuVDsNt+3P5fL5s+45B11hUgXmZsn2lZZk8Mugai7K92fazts/aPmP7o23GqcWyKTgvtVg2tq+3/Q+2X8rn5Q/bjPN221/Nl8sLticGX2kuIvjr8idpTNI/S/ppSeslvSRp27JxHpT0+aprLTAvvyTpDkmvdBh+n6RnJFnSXZJeqLrmHuZlStLXq66z4LzcJumO/P5Nkr7T5jNWi2VTcF5qsWzy9/rG/P51kl6QdNeycX5T0mP5/T2SvlpVvbTQi9kuaS4iLkTEoqSjknZXXNOaRMRzkv5jhVF2S/pyZJ6XdIvt2wZT3eoUmJfaiIhXI+Jb+f3/lHRO0sZlo9Vi2RScl1rI3+v/yh9el/8t35Nkt6Qv5feflrTDtgdU4lsQ6MVslHSx5fG82n9A35f/FH7a9ubBlFa6ovNaF5P5z+VnbP9c1cUUkf9kf4+y1mCr2i2bFeZFqsmysT1m+0VJlyV9IyI6LpeIuCrpiqRbB1tlhkAvz99KmoiId0v6hv7/GxvV+Zay8178gqQ/lfS1iuvpyvaNkv5a0u9ExA+qrqcXXealNssmIq5FxC9K2iRpu+3bq66pEwK9mEuSWlvcm/Ln3hQRr0XEj/KHfy7pzgHVVrau81oXEfGDpZ/LEXFM0nW2N1RcVke2r1MWgH8REX/TZpTaLJtu81K3ZSNJEfF9Sc9K2rls0JvLxfY6STdLem2w1WUI9GJOSdpqe4vt9co2fEy3jrCsL3OXsn7DOpqW9MF8j4q7JF2JiFerLmotbP/UUl+m7e3KPu+VrGjd5HU+LulcRHy2w2i1WDZF5qUuy8b2uO1b8vs3SLpH0j8tG21a0ofy+/dL+mbkW0gHbV0VE62biLhqe5+k48r2eHkiIs7YPiCpGRHTkn7b9i5JV5VtqHuwsoJXYPtJZXsYbLA9L+mTyjb0KCIek3RM2d4Uc5LekPThairtrsC83C/pN2xflfTfkvZUtaIVcLekX5P0j3l/rST9nqR3SLVbNkXmpS7L5jZJX7I9puxL56mI+Pqydf9xSV+xPads3d9TVbEc+g8AiaDLBQASQaADQCIIdABIBIEOAIkg0AEgEQQ6ACSCQAeARPwf/aNBh45s5KwAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jqJOOOLsinEj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}