Module_01_Lab_04_LinearAlgebra.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Harsha-070/FMML_IIITH/blob/main/Module_01_Lab_04_LinearAlgebra.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mU9LZjQxiQT2"
      },
      "source": [
        "# Transforming data using linear algebra\n",
        "\n",
        "FMML Module 1, Lab 4<br>\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "KBe4ZA32UTbv"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from keras.datasets import mnist"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "Rd7LRGvJ3Pt1"
      },
      "outputs": [],
      "source": [
        "#@title\n",
        "\n",
        "from IPython.display import Latex as lt\n",
        "#Plotting functions\n",
        "#(You DON'T need to understand how these functions)\n",
        "\n",
        "# function to plot a grid\n",
        "def plotGrid(transform, unit, linestyle = ':', fig=None, ax=None):\n",
        "  lim1 = -100\n",
        "  lim2 = 100\n",
        "  def mat2xy(start, end):\n",
        "    if len(start.shape)==1:\n",
        "      start = np.expand_dims(start,0)\n",
        "      end = np.expand_dims(end,0)\n",
        "    nan = np.ones(len(start))*np.nan\n",
        "    x = np.stack((start[:,0], end[:,0], nan)).T.reshape(-1)\n",
        "    y = np.stack((start[:,1], end[:,1], nan)).T.reshape(-1)\n",
        "    return x, y\n",
        "\n",
        "  def parallellines(axis, addend, lines, unit):\n",
        "    addend = np.repeat(np.expand_dims(addend,0), lines*2, 0)\n",
        "    unit = np.expand_dims(np.arange(-lines, lines)*unit,1)\n",
        "    unit = unit-lines\n",
        "    addend = addend*unit\n",
        "    lines = np.expand_dims(axis,0) + addend\n",
        "    return np.concatenate((lines, lines*-1))\n",
        "\n",
        "  if fig is None:\n",
        "    fig, ax = plt.subplots(figsize=(10,10))\n",
        "  transform = transform.astype(np.float)\n",
        "  xaxis = transform[0]\n",
        "  yaxis = transform[1]\n",
        "\n",
        "  # plot lines parallel to the x axis\n",
        "  lines1= parallellines(xaxis*lim1, yaxis, 100,unit )\n",
        "  lines2 = parallellines(xaxis*lim2, yaxis, 100,unit )\n",
        "  x,y = mat2xy(lines1, lines2)\n",
        "  plt.plot(x,y, linestyle+'k', linewidth=0.5)\n",
        "  # plot x axis\n",
        "  x,y = mat2xy(xaxis*lim1, xaxis*lim2)\n",
        "  plt.plot(x,y,linestyle, color = '#440077')\n",
        "\n",
        "  # plot  lines parallel to the y axis\n",
        "  lines1= parallellines(yaxis*lim1, xaxis, 100,unit)\n",
        "  lines2 = parallellines(yaxis*lim2, xaxis, 100,unit)\n",
        "  x,y = mat2xy(lines1, lines2)\n",
        "  plt.plot(x,y, linestyle+'k', linewidth=0.5)\n",
        "  # plot y axis\n",
        "  x,y = mat2xy(yaxis*lim1, yaxis*lim2)\n",
        "  plt.plot(x,y,linestyle, color= '#aa5500')\n",
        "\n",
        "  return fig, ax\n",
        "\n",
        "def plotData(X, y, xlabel = 'hole', ylabel = 'bound', fig=None, ax = None):\n",
        "\n",
        "  if fig is None:\n",
        "    fig, ax = plt.subplots()\n",
        "  for ii in range(nclasses):\n",
        "    plt.scatter(X[y==ii,0], X[y==ii, 1])\n",
        "  plt.legend([str(i) for i in range(nclasses)])\n",
        "  plt.xlabel(xlabel)\n",
        "  plt.ylabel(ylabel)\n",
        "  lim2 = X.max()\n",
        "  lim1 = X.min()\n",
        "  add = abs(lim1-lim2)/5\n",
        "  return fig, ax\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "cellView": "form",
        "id": "7mqG-dafVpya"
      },
      "outputs": [],
      "source": [
        "#@title\n",
        "\n",
        "from IPython.display import Latex as lt\n",
        "#Plotting functions\n",
        "#(You DON'T need to understand how these functions)\n",
        "\n",
        "# function to plot a grid\n",
        "def plotGrid(transform, unit, linestyle = ':', fig=None, ax=None):\n",
        "  lim1 = -100\n",
        "  lim2 = 100\n",
        "  def mat2xy(start, end):\n",
        "    if len(start.shape)==1:\n",
        "      start = np.expand_dims(start,0)\n",
        "      end = np.expand_dims(end,0)\n",
        "    nan = np.ones(len(start))*np.nan\n",
        "    x = np.stack((start[:,0], end[:,0], nan)).T.reshape(-1)\n",
        "    y = np.stack((start[:,1], end[:,1], nan)).T.reshape(-1)\n",
        "    return x, y\n",
        "\n",
        "  def parallellines(axis, addend, lines, unit):\n",
        "    addend = np.repeat(np.expand_dims(addend,0), lines*2, 0)\n",
        "    unit = np.expand_dims(np.arange(-lines, lines)*unit,1)\n",
        "    unit = unit-lines\n",
        "    addend = addend*unit\n",
        "    lines = np.expand_dims(axis,0) + addend\n",
        "    return np.concatenate((lines, lines*-1))\n",
        "\n",
        "  if fig is None:\n",
        "    fig, ax = plt.subplots(figsize=(10,10))\n",
        "  transform = transform.astype(np.float)\n",
        "  xaxis = transform[0]\n",
        "  yaxis = transform[1]\n",
        "\n",
        "  # plot lines parallel to the x axis\n",
        "  lines1= parallellines(xaxis*lim1, yaxis, 100,unit )\n",
        "  lines2 = parallellines(xaxis*lim2, yaxis, 100,unit )\n",
        "  x,y = mat2xy(lines1, lines2)\n",
        "  plt.plot(x,y, linestyle+'k', linewidth=0.5)\n",
        "  # plot x axis\n",
        "  x,y = mat2xy(xaxis*lim1, xaxis*lim2)\n",
        "  plt.plot(x,y,linestyle, color = '#440077')\n",
        "\n",
        "  # plot  lines parallel to the y axis\n",
        "  lines1= parallellines(yaxis*lim1, xaxis, 100,unit)\n",
        "  lines2 = parallellines(yaxis*lim2, xaxis, 100,unit)\n",
        "  x,y = mat2xy(lines1, lines2)\n",
        "  plt.plot(x,y, linestyle+'k', linewidth=0.5)\n",
        "  # plot y axis\n",
        "  x,y = mat2xy(yaxis*lim1, yaxis*lim2)\n",
        "  plt.plot(x,y,linestyle, color= '#aa5500')\n",
        "\n",
        "  return fig, ax\n",
        "\n",
        "def plotData(X, y, xlabel = 'hole', ylabel = 'bound', fig=None, ax = None):\n",
        "\n",
        "  if fig is None:\n",
        "    fig, ax = plt.subplots()\n",
        "  for ii in range(nclasses):\n",
        "    plt.scatter(X[y==ii,0], X[y==ii, 1])\n",
        "  plt.legend([str(i) for i in range(nclasses)])\n",
        "  plt.xlabel(xlabel)\n",
        "  plt.ylabel(ylabel)\n",
        "  lim2 = X.max()\n",
        "  lim1 = X.min()\n",
        "  add = abs(lim1-lim2)/5\n",
        "  return fig, ax\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9d5dmEXxaktp"
      },
      "source": [
        "# Matrix transformations on data\n",
        "\n",
        "A 2D coordinate system is defined by its basis vectors, i and j. In the standard coordinate system (Let us call it T0), the basis vectors are\n",
        "\n",
        "$$\\begin{equation}\n",
        "i = \\left\\{  \\begin{aligned}1 \\\\ 0 \\end{aligned} \\right\\}\n",
        "\\end{equation}$$\n",
        "and\n",
        "$$\\begin{equation} j = \\left\\{ \\begin{aligned} 0 \\\\ 1\\end{aligned} \\right\\} \\end{equation}$$\n",
        "\n",
        "We can use any two vectors as basis vectors for a new coordinate system as long as they are not colinear. For example, let us call this new coordinate system T1:\n",
        "\n",
        "$$\\begin{equation}\n",
        "i = \\left\\{  \\begin{aligned}1 \\\\ -1 \\end{aligned} \\right\\}\n",
        "\\end{equation}$$\n",
        "and\n",
        "$$\\begin{equation} j = \\left\\{ \\begin{aligned} 0 \\\\ 2 \\end{aligned} \\right\\} \\end{equation}$$\n",
        "\n",
        "Suppose we have a point [a,b] in the T1 coordinate system. Its representation in the standard system T0 can be obtained by the following matrix multiplication:\n",
        "\n",
        "$$ \\begin{equation}\n",
        "\\left\\{  \\begin{aligned}a' \\\\ b' \\end{aligned} \\right\\} =\n",
        "\\left\\{  \\begin{aligned}&1 & 0 \\\\ -&1 & 2 \\end{aligned} \\right\\}\n",
        "\\left\\{  \\begin{aligned}a \\\\ b \\end{aligned} \\right\\}\n",
        "\\end{equation}$$\n",
        "where the columns of the matrix are the basis vectors of T1.\n",
        "\n",
        "Let us see this in action:\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K8SCFKqfbM-l",
        "outputId": "b9ea644a-b959-4dc1-8e25-31f6e1f2fe95"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Data in T0 =  [5 3]\n",
            "Data in T1 =  [5 4]\n"
          ]
        }
      ],
      "source": [
        "T0 = np.array([[1,0],[0,1]])\n",
        "T1 = np.array([[1,0], [-1,2]])\n",
        "\n",
        "data1 = np.array([5,4]) # the data in T1 coordinate system\n",
        "data0 = np.matmul(T1, data1) # the data in T0 coordinate system\n",
        "\n",
        "print('Data in T0 = ', data0)\n",
        "print('Data in T1 = ', data1)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kIfEWZ1RQeve"
      },
      "source": [
        "We can visualize this below. T0 is shown with dotted lines and T1 is shown with solid lines."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 879
        },
        "id": "jSuTaeSDQoYK",
        "outputId": "4fdc4a95-fec0-4e86-e1f8-ff10ce2eb690"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-5-eb238eeb2a58>:30: DeprecationWarning: `np.float` is a deprecated alias for the builtin `float`. To silence this warning, use `float` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.float64` here.\n",
            "Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations\n",
            "  transform = transform.astype(np.float)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x1000 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "fig, ax = plotGrid(T1.T, 1,'-') # custom plotting function, no need to understand this\n",
        "plotGrid(T0.T, 1, fig=fig, ax=ax) # custom plotting function, no need to understand this\n",
        "plt.scatter(data0[0], data0[1])\n",
        "ax.set_xlim(-10,10)\n",
        "ax.set_ylim(-10,10)\n",
        "ax.set_xticks([]);\n",
        "ax.set_yticks([]);\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nUb3oYWIgdya"
      },
      "source": [
        "Look at the coordinates of the blue dot. In T0 (dotted lines), the position is [5,3] where it is [5,4] in T1. Feel free to experiment with different data points and coordinate systems.\n",
        "\n",
        "Remember that we can achieve the same thing by post-multiplying the transpose of the transformation matrix to the data. This will come in handy when transforming multiple data points at once:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SD4pqOMndf-4",
        "outputId": "81b4faf5-09d9-451e-ac94-ae82b3b48592"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[5 3]\n",
            "[5 3]\n"
          ]
        }
      ],
      "source": [
        "data0_a = np.matmul(T1, data1)\n",
        "data0_b = np.matmul(data1, T1.T)\n",
        "print(data0_a)\n",
        "print(data0_b)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z-Kui3dSESPm"
      },
      "source": [
        "Why is transforming data useful? Data transformations cause the distance between data points to change. This will affect distance-based algorithms such as nearest neighbour"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nE0NbpYIC9ou",
        "outputId": "2217f4db-acce-4bd4-bddd-6e7bc1ec21a8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Distance between A and B in T1 =  8.06225774829855\n",
            "Distance between B and C in T1 =  4.123105625617661\n",
            "Distance between A and C in T1 =  4.47213595499958\n",
            "\n",
            "Distance between A and B in T0 =  15.033296378372908\n",
            "Distance between B and C in T0 =  9.055385138137417\n",
            "Distance between A and C in T0 =  6.324555320336759\n"
          ]
        }
      ],
      "source": [
        "# let us define 3 points in T1\n",
        "A1 = np.array([3,3])\n",
        "B1 = np.array([2,-5])\n",
        "C1 = np.array([1,-1])\n",
        "\n",
        "# the corresponding points in T0:\n",
        "A0 = np.matmul(T1, A1)\n",
        "B0 = np.matmul(T1, B1)\n",
        "C0 = np.matmul(T1, C1)\n",
        "\n",
        "# function to calculate Euclidean distance:\n",
        "def dist(a, b):\n",
        "  diff = a-b\n",
        "  sq = diff*diff\n",
        "  return np.sqrt(sq.sum())\n",
        "\n",
        "# distance between the points in T1\n",
        "print('Distance between A and B in T1 = ', dist(A1, B1))\n",
        "print('Distance between B and C in T1 = ', dist(B1, C1))\n",
        "print('Distance between A and C in T1 = ', dist(A1, C1))\n",
        "\n",
        "print('')\n",
        "# distnace between the points in T0\n",
        "print('Distance between A and B in T0 = ', dist(A0, B0))\n",
        "print('Distance between B and C in T0 = ', dist(B0, C0))\n",
        "print('Distance between A and C in T0 = ', dist(A0, C0))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YE95JMniUtDm"
      },
      "source": [
        "We see that in T1, B and C are the closest whereas in T0, A and C are the closest. These kinds of changes will affect the predictions returned by the nearest neighbour algorithm."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PFbpTSBdV29C"
      },
      "source": [
        "# Transformations on MNIST\n",
        "\n",
        "Let us experiment with a subset of the MNIST dataset. We will extract two features from the database for our experiment."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kMvUzC2GUawS"
      },
      "source": [
        "Functions for nearest neighbour, accuracy and feature extraction. (from previous labs)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "22ayl0sViF5-"
      },
      "outputs": [],
      "source": [
        "#@title\n",
        "\n",
        "def NN1(traindata, trainlabel, query):\n",
        "  diff  = traindata - query  # find the difference between features. Numpy automatically takes care of the size here\n",
        "  sq = diff*diff # square the differences\n",
        "  dist = sq.sum(1) # add up the squares\n",
        "  label = trainlabel[np.argmin(dist)] # our predicted label is the label of the training data which has the least distance from the query\n",
        "  return label\n",
        "\n",
        "def NN(traindata, trainlabel, testdata):\n",
        "  # we will run nearest neighbour for each sample in the test data\n",
        "  # and collect the predicted classes in an array using list comprehension\n",
        "  predlabel = np.array([NN1(traindata, trainlabel, i) for i in testdata])\n",
        "  return predlabel\n",
        "\n",
        "def Accuracy(gtlabel, predlabel):\n",
        "  assert len(gtlabel)==len(predlabel), \"Length of the groundtruth labels and predicted labels should be the same\"\n",
        "  correct = (gtlabel==predlabel).sum() # count the number of times the groundtruth label is equal to the predicted label.\n",
        "  return correct/len(gtlabel)\n",
        "\n",
        "def cumArray(img):\n",
        "  img2 = img.copy()\n",
        "  for ii in range(1, img2.shape[1]):\n",
        "    img2[ii,:] = img2[ii,:] + img2[ii-1,:]  # for every row, add up all the rows above it.\n",
        "  img2 = img2>0\n",
        "  return img2\n",
        "\n",
        "def getHolePixels(img):\n",
        "  im1 = cumArray(img)\n",
        "  im2 = np.rot90(cumArray(np.rot90(img)), 3) # rotate and cumulate it again for differnt direction\n",
        "  im3 = np.rot90(cumArray(np.rot90(img, 2)), 2)\n",
        "  im4 = np.rot90(cumArray(np.rot90(img, 3)), 1)\n",
        "  hull =  im1 & im2 & im3 & im4 # this will create a binary image with all the holes filled in.\n",
        "  hole = hull & ~ (img>0) # remove the original digit to leave behind the holes\n",
        "  return hole\n",
        "\n",
        "def getHullPixels(img):\n",
        "  im1 = cumArray(img)\n",
        "  im2 = np.rot90(cumArray(np.rot90(img)), 3) # rotate and cumulate it again for differnt direction\n",
        "  im3 = np.rot90(cumArray(np.rot90(img, 2)), 2)\n",
        "  im4 = np.rot90(cumArray(np.rot90(img, 3)), 1)\n",
        "  hull =  im1 & im2 & im3 & im4 # this will create a binary image with all the holes filled in.\n",
        "  return hull\n",
        "\n",
        "def minus(a, b):\n",
        "  return a & ~ b\n",
        "\n",
        "def getBoundaryPixels(img):\n",
        "  img = img.copy()>0  # binarize the image\n",
        "  rshift = np.roll(img, 1, 1)\n",
        "  lshift = np.roll(img, -1 ,1)\n",
        "  ushift = np.roll(img, -1, 0)\n",
        "  dshift = np.roll(img, 1, 0)\n",
        "  boundary = minus(img, rshift) | minus(img, lshift) | minus(img, ushift) | minus(img, dshift)\n",
        "  return boundary\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RzF-9mGxUkWC"
      },
      "source": [
        "Get the MNIST dataset:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bHz5BVmLUjzb",
        "outputId": "c92fdef1-dc6b-4589-e0e6-d3d8198c6abb"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11490434/11490434 [==============================] - 0s 0us/step\n"
          ]
        }
      ],
      "source": [
        "#loading the dataset\n",
        "(train_X, train_y), (test_X, test_y) = mnist.load_data()\n",
        "train_X = train_X/255\n",
        "test_X = test_X/255\n",
        "\n",
        "nclasses = 4\n",
        "\n",
        "# get only for the first 4 classes\n",
        "train_X = train_X[train_y<nclasses]\n",
        "train_y = train_y[train_y<nclasses]\n",
        "test_X = test_X[test_y<nclasses]\n",
        "test_y = test_y[test_y<nclasses]\n",
        "\n",
        "train_X = train_X[::100].copy() # We are only taking a subset of the training set\n",
        "train_y = train_y[::100].copy() # do the same to the labels\n",
        "test_X = test_X[::100].copy() # taking a subset of the test set. This code takes every 500th sample\n",
        "test_y = test_y[::100].copy()\n",
        "\n",
        "# get all the features\n",
        "train_hole = np.array([getHolePixels(i).sum() for i in train_X])\n",
        "test_hole = np.array([getHolePixels(i).sum() for i in test_X])\n",
        "train_bound = np.array([getBoundaryPixels(i).sum() for i in train_X])\n",
        "test_bound = np.array([getBoundaryPixels(i).sum() for i in test_X])\n",
        "# train_hull = np.array([getHullPixels(i).sum() for i in train_X])\n",
        "# test_hull = np.array([getHullPixels(i).sum() for i in test_X])\n",
        "# train_sum = np.sum(train_X, (1,2))/(28*28)\n",
        "# test_sum = np.sum(test_X, (1,2))/(28*28)\n",
        "\n",
        "# create the train and test set by combining the appropriate features\n",
        "train_feats = np.vstack((train_hole,train_bound)).transpose()\n",
        "test_feats = np.vstack((test_hole, test_bound)).transpose()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I7qGVrlnQCUy"
      },
      "source": [
        "Let us plot the samples and see what they look like:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "saRzHfi9QAGd",
        "outputId": "698f83f1-8b56-4666-8967-6e24ddff64f4"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(-100.0, 300.0)"
            ]
          },
          "metadata": {},
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# fix limits of x and y axis so that we can see what is going on\n",
        "xlim=[-100,300]\n",
        "ylim=[-100,300]\n",
        "fig, ax = plotData(train_feats, train_y)\n",
        "ax.set_xlim(xlim)\n",
        "ax.set_ylim(ylim)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hCb4DikQP1ck"
      },
      "source": [
        "Check the baseline accuracy on the test set:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rxVr6bd9PzlI",
        "outputId": "b9e80e63-e992-41b0-f216-13b5baab965b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Baseline accuracy =  0.7619047619047619\n"
          ]
        }
      ],
      "source": [
        "test_pred = NN(train_feats, train_y, test_feats)\n",
        "acc = Accuracy(test_y, test_pred)\n",
        "print('Baseline accuracy = ', acc)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nl8Noo8pZRek"
      },
      "source": [
        "Let us try transforming the features and checking their accuracy:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "8dUPWRsEZKwo"
      },
      "outputs": [],
      "source": [
        "transform = np.array([[0.5,-0.5],[0,2.5]])\n",
        "\n",
        "train_feats_t = np.matmul(train_feats, transform)\n",
        "test_feats_t = np.matmul(test_feats, transform)  # whatever transform we are applying to the training set should be applied to the test set also"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VqMBUAfqIUcC",
        "outputId": "d2e2eea5-18e1-430e-efe2-e6483271ced0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[ 0.5 -0.5]\n",
            " [ 0.   2.5]]\n"
          ]
        }
      ],
      "source": [
        "print(transform)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "iRH4VckHZaWv",
        "outputId": "73998c9b-4f29-4e2f-8489-adf7de030ec6"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(-100.0, 300.0)"
            ]
          },
          "metadata": {},
          "execution_count": 16
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "fig, ax = plotData(train_feats_t, train_y)\n",
        "ax.set_xlim(xlim)\n",
        "ax.set_ylim(ylim)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "l9fknczfZdYF",
        "outputId": "d77df52e-2310-4f73-c012-44766ebdf0f5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy after transform =  0.8095238095238095\n"
          ]
        }
      ],
      "source": [
        "test_pred = NN(train_feats_t, train_y, test_feats_t)\n",
        "acc = Accuracy(test_y, test_pred)\n",
        "print('Accuracy after transform = ', acc)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OFBQlAnNZ3on"
      },
      "source": [
        "## Questions:\n",
        "1. Experiment with different transformation matrices and check the accuracy\n",
        "2. Will the same transform used for these two features also work for other features?"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#1.Experiment with different transformation matrices and check their accuracy\n",
        "\n",
        " On what kind of transformation you are interested in (e.g., translation, rotation, scaling, etc.) and what you're trying to achieve (e.g., image processing, 3D graphics, data augmentation, etc.).\n",
        "\n",
        " with a general outline of how you might approach such an experiment:\n",
        "\n",
        "1. **Define the Transformation:**\n",
        "\n",
        "First, you need to decide on the type of transformation you want to experiment with. Let's take a simple example of a 2D rotation matrix as follows:\n",
        "\n",
        "    \n",
        "    R(theta) = | cos(theta)  -sin(theta) |\n",
        "               | sin(theta)   cos(theta) |\n",
        "  \n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "  You can replace this with any other transformation matrix depending on your specific experiment.\n",
        "\n",
        "   2. **Generate Data:**\n",
        "\n",
        " Create a dataset or sample data points that you want to transform. For instance, you can create a set of 2D points.\n",
        "\n",
        "3. **Apply Transformation:**\n",
        "\n",
        " Apply the transformation matrix to each data point in your dataset. For a rotation matrix, you'd multiply each point's coordinates by the matrix.\n",
        "\n",
        "4. **Evaluate Accuracy:**\n",
        "\n",
        " If you have a specific task in mind, such as image classification or object recognition, you can evaluate the accuracy of your model before and after applying the transformation. For example, you might want to rotate images of handwritten digits and see how it affects the accuracy of a digit recognition model.\n",
        "\n",
        "5. **Repeat and Experiment:**\n",
        "\n",
        " You can repeat the above steps with different transformation parameters (e.g., varying the rotation angle) or using different transformation matrices to see how they affect the accuracy of your model.\n",
        "\n",
        "Here's a simplified Python example using numpy to demonstrate a rotation transformation:\n",
        "\n",
        "```python\n",
        "import numpy as np\n",
        "\n",
        "# Define rotation matrix\n",
        "theta = np.radians(30)  # 30 degrees\n",
        "rotation_matrix = np.array([[np.cos(theta), -np.sin(theta)],\n",
        "                            [np.sin(theta), np.cos(theta)]])\n",
        "\n",
        "# Generate sample data\n",
        "data_points = np.array([[1, 0],\n",
        "                        [0, 1],\n",
        "                        [-1, 0],\n",
        "                        [0, -1]])\n",
        "\n",
        "# Apply the rotation to each data point\n",
        "transformed_data = np.dot(data_points, rotation_matrix)\n",
        "\n",
        "# Print the transformed data\n",
        "print(transformed_data)\n",
        "```\n",
        "\n",
        "This code rotates a set of 2D points by 30 degrees. You can adapt this code to your specific experiment and accuracy evaluation needs."
      ],
      "metadata": {
        "id": "4iTF4yAPJAyY"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#2.Will the same transform used for these two features also work for other features?\n",
        "\n",
        "The transform used for one set of features may or may not work for other features, depending on the nature of the features and the specific transformation applied. Here are some factors to consider:\n",
        "\n",
        "1. **Similarity of Features:**\n",
        "\n",
        " If the features you want to transform are similar in nature and have the same underlying characteristics, there's a higher chance that the same transformation will work for both sets of features. For example, if you have two sets of images with similar content but different orientations, a rotation transformation that works for one set of images may work for the other.\n",
        "\n",
        "2. **Dimensionality:**\n",
        "\n",
        " The dimensionality of the features is crucial. A transformation that works for 2D data may not work for 3D data without modification. You may need to adapt the transformation matrix and approach to accommodate the different dimensionality.\n",
        "\n",
        "3. **Type of Transformation:**\n",
        "\n",
        " The type of transformation matters. Some transformations are more general and applicable across various types of data, such as translation and scaling. Others, like rotation or perspective transformation, may be more specific and depend on the geometry and characteristics of the data.\n",
        "\n",
        "4. **Domain-specific Considerations:**\n",
        "\n",
        " Depending on the application domain, you may need to consider domain-specific factors. For example, in computer vision, different objects and scenes may require different transformations for accurate processing.\n",
        "\n",
        "5. **Data Preprocessing:**\n",
        "\n",
        " Sometimes, it's necessary to preprocess the features differently before applying the same transformation. For instance, if one set of features has a different scale or origin point than the other, you may need to normalize them first.\n",
        "\n",
        "6. **Transformation Parameters:**\n",
        "\n",
        " Even for the same type of transformation, the parameters may vary. For example, if you're rotating images, the rotation angle may need to be adjusted for different sets of images.\n",
        "\n",
        "\n",
        "The same transformation will work for other features depends on the nature of the features, the type of transformation, and any domain-specific considerations. It's often necessary to experiment and adapt the transformation to the specific characteristics of the features you're working with."
      ],
      "metadata": {
        "id": "leIdF2AaKKqZ"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "36TOA47xak20"
      },
      "source": [
        "# Data normalization\n",
        "\n",
        "Sometimes the features of our data have vastly different scales. This will cause the learning algorithm to give more importance to certain features, reducing its performance. Data normalization is a method in which we transform the features so that they have similar scales.\n",
        "\n",
        "Three commonly used feature scaling techniques are rescaling, mean normalization and z-score normalization. Here, we will talk about the simplest one: rescaling.\n",
        "\n",
        "$$\\begin{equation}\n",
        "x' = \\frac {x -min(x)} { max(x) - min(x)}\n",
        "\\end{equation}$$\n",
        "\n",
        "\n",
        "\n",
        "For more information, see [this page](https://towardsdatascience.com/data-normalization-in-machine-learning-395fdec69d02)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "ni19QKDLZzeo"
      },
      "outputs": [],
      "source": [
        "def rescale(data):\n",
        "  return (data - data.min())/(data.max() - data.min())"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8k83ZMtKeZrQ"
      },
      "source": [
        "We have to apply the rescaling to each feature individually. Also remember to apply the same transform we are using on the train set to the test set."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "1JLOPwhvehpR"
      },
      "outputs": [],
      "source": [
        "train_feats_rescaled_x = rescale(train_feats[:,0])\n",
        "train_feats_rescaled_y = rescale(train_feats[:,1])\n",
        "train_feats_rescaled = np.stack((train_feats_rescaled_x, train_feats_rescaled_y),1)\n",
        "\n",
        "test_feats_rescaled_x = rescale(test_feats[:,0])\n",
        "test_feats_rescaled_y = rescale(test_feats[:,1])\n",
        "test_feats_rescaled = np.stack((test_feats_rescaled_x, test_feats_rescaled_y),1)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RaZVy9vxfKwX"
      },
      "source": [
        "Let us plot the rescaled features:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "LXyatpwZevOH",
        "outputId": "f57f97c4-991d-4ba4-e8dc-cf68c6198bfa"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "fig, ax = plotData(train_feats_rescaled, train_y)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZjBgxE6zglsL"
      },
      "source": [
        "This type of rescaling makes all the features between 0 and 1.\n",
        "\n",
        "Let us calculate the accuracy obtained by this transform:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "hKQnsj-KgNZc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "84b680b3-ee55-4e00-a55c-9efab3c1b589"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy after transform =  0.8095238095238095\n"
          ]
        }
      ],
      "source": [
        "test_pred = NN(train_feats_rescaled, train_y, test_feats_rescaled)\n",
        "acc = Accuracy(test_y, test_pred)\n",
        "print('Accuracy after transform = ', acc)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7qYX90Gqg-jO"
      },
      "source": [
        "All 2D linear transformations can be repreented by a transformation matrix. So what is the matrix associated with the rescaling function? Actually, we cannot represent rescaling with a matrix multiplication, because it is not a linear transform. Rescaling involves shifting the origin of the data, which is not allowed under linear transformations.\n",
        "\n",
        "We can represent rescaling as a matrix multiplication followed by a vector addition. Let our first feature vector be called X and second feature vector be called Y. Suppose we want to rescale a data point [a,b]\n",
        "\n",
        "$$ \\begin{equation}\n",
        " \\left\\{  \\begin{aligned}a' \\\\ b' \\end{aligned} \\right\\} =\n",
        " \\left\\{  \\begin{aligned} \\frac{a - min(X)}{max(X) - min(X)} \\\\ \\frac{b - min(Y)}{max(Y) - min(Y)} \\end{aligned} \\right\\} =\n",
        " \\left\\{  \\begin{aligned}&\\frac{1}{max(X)-min(X)} &0\\\\ &0 &\\frac{1}{max(Y)-min(Y)} \\end{aligned}\n",
        " \\right\\}\\left\\{  \\begin{aligned}a \\\\ b \\end{aligned} \\right\\} +\n",
        " \\left\\{  \\begin{aligned} \\frac{ -min(X)}{max(X) - min(X)} \\\\ \\frac{-min(Y)}{max(Y) - min(Y)} \\end{aligned} \\right\\}\n",
        "\\end{equation}$$\n",
        "\n",
        "You can verify this yourself if you wish, though it is not necessary.\n",
        "\n"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}