2_1_Supervised_Learning.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "mount_file_id": "https://github.com/jamalromero/colab/blob/main/udl_ch2.ipynb",
      "authorship_tag": "ABX9TyNxsp5KVBZH84bD+0ZRaR8N",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "jbang-ijava",
      "display_name": "java (IJava/j!)"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/jamalromero/colab/blob/main/2_1_Supervised_Learning.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Understanding Deep Learning [book](https://udlbook.github.io/udlbook/)\n",
        "## Notebook 2.1 Supervised Learning"
      ],
      "metadata": {
        "id": "xIr_Md1WRvoi"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "id": "pjUM1cc8pP04"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Keep Python runtime and install java"
      ],
      "metadata": {
        "id": "-PY9c_q42J1S"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tGjdJncAIbnT"
      },
      "outputs": [],
      "source": [
        "!apt install openjdk-17-jdk"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Install Juyter Java kernel"
      ],
      "metadata": {
        "id": "VTvVS96Q2Tzd"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install jbang\n",
        "import jbang\n",
        "jbang.exec(\"trust add https://github.com/jupyter-java\")\n",
        "jbang.exec(\"install-kernel@jupyter-java\")"
      ],
      "metadata": {
        "id": "advvIMn1Nn0e"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Change Runtime to Java kernel and test Java runtime."
      ],
      "metadata": {
        "id": "ro6LZ1iT2cig"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "System.getProperty(\"java.version\")"
      ],
      "metadata": {
        "id": "cwZaX7xHNcKk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Load external script to include maven dependency, import packages and intial startup script. Example:\n",
        "\n",
        "```\n",
        "%%loadFromPOM\n",
        "<dependency>\n",
        "    <groupId>tech.tablesaw</groupId>\n",
        "    <artifactId>tablesaw-core</artifactId>\n",
        "    <version>0.43.1</version>\n",
        "</dependency>\n",
        "\n",
        "import java.util.stream.*;\n",
        "\n",
        "System.getProperty(\"java.version\");\n",
        "\n",
        "```\n",
        "\n"
      ],
      "metadata": {
        "id": "z6jlXMb82p9i"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%load /content/drive/MyDrive/colab/init_script.ipynb"
      ],
      "metadata": {
        "id": "HYw6JGoL4A29"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "NDManager manager = NDManager.newBaseManager();"
      ],
      "metadata": {
        "id": "RpqXcCwcSyfa"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Supervised Learning\n",
        "The purpose of this notebook is to explore the linear regression model discussed in Chapter 2 of the book.\n",
        "\n"
      ],
      "metadata": {
        "id": "wXCM4gc6ne-M"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "// Create some input data\n",
        "var x = manager.create(new double[]{0.03, 0.19, 0.34, 0.46, 0.78, 0.81, 1.08, 1.18, 1.39, 1.60, 1.65, 1.90});\n",
        "var y = manager.create(new double[]{0.67, 0.85, 1.05, 1.0, 1.40, 1.5, 1.3, 1.54, 1.55, 1.68, 1.73, 1.6})"
      ],
      "metadata": {
        "id": "kRLpyfzGTTmC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Define 1D linear regression model\n",
        "Equation: $y=\\phi_0 + \\phi_1x$"
      ],
      "metadata": {
        "id": "nPzrhzTpCwVU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "NDArray f1D(NDArray x, double phi0, double phi1) {\n",
        "  return x.mul(phi1).add(phi0);\n",
        "}"
      ],
      "metadata": {
        "id": "T1LGYBPnT2Mt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Function to help plot the data"
      ],
      "metadata": {
        "id": "1Tn4_QX7EHi1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "Canvas plot(NDArray x, NDArray y, double phi0, double phi1) {\n",
        "  double[][] points = toDoubleArray(x.stack(y, 1));\n",
        "  var canvas = ScatterPlot.of(points, '*', Color.BLUE).canvas();\n",
        "  points = toDoubleArray(x.stack(f1D(x, phi0, phi1), 1));\n",
        "  canvas.add(LinePlot.of(points, Color.RED));\n",
        "  return canvas;\n",
        "}"
      ],
      "metadata": {
        "id": "Jv5M-11RUq0S"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "var canvas = plot(x, y, 0.84, 0.5);\n",
        "canvas.setBound(new double[]{0f, 0f}, new double[]{2f, 2f});\n",
        "canvas.setAxisLabels(\"Input X\", \"Output Y\");\n",
        "canvas.toBufferedImage(800, 600)"
      ],
      "metadata": {
        "id": "J0zr9tfgWF51",
        "outputId": "32f4cfbf-5f22-48ef-b494-345e252cc5cd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 617
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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",
            "text/plain": [
              "BufferedImage@25da64c5: type = 2 DirectColorModel: rmask=ff0000 gmask=ff00 bmask=ff amask=ff000000 IntegerInterleavedRaster: width = 800 height = 600 #Bands = 4 xOff = 0 yOff = 0 dataOffset[0] 0"
            ]
          },
          "metadata": {},
          "execution_count": 177
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Function to calculate the loss"
      ],
      "metadata": {
        "id": "5VMWr2-qMbdt"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "double computeLoss(NDArray x, NDArray y, double phi0, double phi1) {\n",
        "    var y_hat = f1D(x, phi0, phi1);\n",
        "    double loss =  y.sub(y_hat).pow(2).sum().toDoubleArray()[0];\n",
        "    return loss;\n",
        "}"
      ],
      "metadata": {
        "id": "qDw1NH3BWVgr"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Change the parameters manually to fit the model\n",
        "\n",
        "First fix $\\phi_1$ and try changing $\\phi_0$ until you can't make the loss go down any more\n",
        "\n",
        "Then fix $\\phi_0$ and try changing $\\phi_1$ until you can't make the loss go down any more\n",
        "\n",
        "Repeat this process until you find a set of parameters that fit the model as in figure 2.2d\n",
        "\n",
        "You can either do this by hand, or if you want to get fancy, write code to descent automatically in this way\n",
        "\n",
        "Start at these values: $\\phi_0 = 0.75$ and $\\phi_1 = 0.54$"
      ],
      "metadata": {
        "id": "89pCwxZ-lFOm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "double phi0 = 0.75; double phi1 = 0.54;\n",
        "var loss = computeLoss(x, y, phi0, phi1);\n",
        "System.out.printf(\"Calculated Loss is %,.2f for phi0 = %,.2f, phi1 = %,.2f %n\", loss, phi0, phi1);\n",
        "phi0 = 1.6; phi1 = -0.8;\n",
        "loss = computeLoss(x, y, phi0, phi1);\n",
        "System.out.printf(\"Calculated Loss is %,.2f for phi0 = %,.2f, phi1 = %,.2f %n\", loss, phi0, phi1);\n",
        "phi0 = 0.84; phi1 = 0.5;\n",
        "loss = computeLoss(x, y, phi0, phi1);\n",
        "System.out.printf(\"Calculated Loss is %,.2f for phi0 = %,.2f, phi1 = %,.2f %n\", loss, phi0, phi1);"
      ],
      "metadata": {
        "id": "BPvaXNp2bhrC",
        "outputId": "dcb21ba6-2953-482d-b288-7af74ce9c7aa",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Calculated Loss is 0.24 for phi0 = 0.75, phi1 = 0.54 \n",
            "Calculated Loss is 10.28 for phi0 = 1.60, phi1 = -0.80 \n",
            "Calculated Loss is 0.20 for phi0 = 0.84, phi1 = 0.50 \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "java.io.PrintStream@63cbd7e3"
            ]
          },
          "metadata": {},
          "execution_count": 182
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "phi0 = 0.77; phi1 = 0.57;\n",
        "loss = computeLoss(x, y, phi0, phi1);\n",
        "System.out.printf(\"Calculated Loss is %,.2f for phi0 = %,.2f, phi1 = %,.2f %n\", loss, phi0, phi1);"
      ],
      "metadata": {
        "id": "tawpzbwhG6gQ",
        "outputId": "92a60cea-5626-4dce-aeda-d2da5916605a",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Calculated Loss is 0.21 for phi0 = 0.77, phi1 = 0.57 \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "java.io.PrintStream@63cbd7e3"
            ]
          },
          "metadata": {},
          "execution_count": 184
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "phi0 = 0.4 ; phi1 = 0.2;\n",
        "var initial_loss = computeLoss(x,y,phi0,phi1);\n",
        "System.out.printf(\"Calculated initial Loss is %,.2f for phi0 = %,.2f, phi1 = %,.2f %n\", initial_loss, phi0, phi1);\n",
        "double learning_rate = 0.01;\n",
        "List<Double> losses = new ArrayList<>();\n",
        "for(int i = 0; i < 200; i++) {\n",
        "  var loss = computeLoss(x,y,phi0 + learning_rate, phi1 + learning_rate);\n",
        "  if(loss < initial_loss) {\n",
        "    phi0 += learning_rate;\n",
        "    phi1 += learning_rate;\n",
        "    initial_loss = loss;\n",
        "    losses.add(loss);\n",
        "  }\n",
        "}\n",
        "var minimum_loss = losses.stream().min(Double::compareTo).get();\n",
        "System.out.printf(\"Calculated Loss is %,.2f for phi0 = %,.2f, phi1 = %,.2f %n\", minimum_loss, phi0, phi1);"
      ],
      "metadata": {
        "id": "CM7N270cOJUL",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f9cf3678-2dd0-496a-ba3c-564587b31ace"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Calculated initial Loss is 7.07 for phi0 = 0.40, phi1 = 0.20 \n",
            "Calculated Loss is 0.21 for phi0 = 0.77, phi1 = 0.57 \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "java.io.PrintStream@63cbd7e3"
            ]
          },
          "metadata": {},
          "execution_count": 186
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "var canvas = plot(x, y, 0.77, 0.55);\n",
        "canvas.setBound(new double[]{0f, 0f}, new double[]{2f, 2f});\n",
        "canvas.setAxisLabels(\"Input X\", \"Output Y\");\n",
        "canvas.setTitle(\"Calculated minimum loss is 0.21 for phi0 = 0.77, phi1 = 0.57 \");\n",
        "canvas.toBufferedImage(800, 600)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 617
        },
        "id": "tsb4XaFRXRh4",
        "outputId": "84977c56-8d20-4592-ce08-2c44a4a277bd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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",
            "text/plain": [
              "BufferedImage@3b5819ef: type = 2 DirectColorModel: rmask=ff0000 gmask=ff00 bmask=ff amask=ff000000 IntegerInterleavedRaster: width = 800 height = 600 #Bands = 4 xOff = 0 yOff = 0 dataOffset[0] 0"
            ]
          },
          "metadata": {},
          "execution_count": 188
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Visualizing the loss function\n",
        "The above process is equivalent to descending coordinate wise on the loss function\n",
        "Now let's plot that function"
      ],
      "metadata": {
        "id": "6NLopMMWns-r"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "var phi0 = manager.arange(0f, 2f, 0.02f, DataType.FLOAT64);\n",
        "var phi1 = manager.arange(-1f, 1f, 0.02f, DataType.FLOAT64);\n",
        "var gridMap = meshgrid(phi0, phi1);\n",
        "var phi0_mesh = gridMap.get(\"x1\");\n",
        "var phi1_mesh  = gridMap.get(\"x2\");\n",
        "var allLosses = manager.zeros(new Shape(phi0.size(), phi1.size()));\n",
        "for(int i = 0; i < phi0.size(); i++) {\n",
        "  for(int j = 0; j < phi1.size(); j++) {\n",
        "     allLosses.set(new NDIndex(i,j), computeLoss(x, y, phi0_mesh.get(i, j).getDouble(), phi1_mesh.get(i,j).getDouble()));\n",
        "  }\n",
        "}\n"
      ],
      "metadata": {
        "id": "T-t98t8EdLYL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "double[][] data = toDoubleArray(allLosses.toType(DataType.FLOAT64, false));\n",
        "var canvas = Heatmap.of(phi0.toDoubleArray(), phi1.toDoubleArray(), data, 256).canvas();\n",
        "//canvas.add(Contour.of(phi0.toDoubleArray(), phi1.toDoubleArray(),data));\n",
        "canvas.setAxisLabels(\"phi0\", \"phi1\");\n",
        "canvas.toBufferedImage(800, 600)"
      ],
      "metadata": {
        "id": "dtQXFpdmEJPe",
        "outputId": "907e66a4-36f0-4a6f-decc-3f6497f5bc1a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 617
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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",
            "text/plain": [
              "BufferedImage@6b2da033: type = 2 DirectColorModel: rmask=ff0000 gmask=ff00 bmask=ff amask=ff000000 IntegerInterleavedRaster: width = 800 height = 600 #Bands = 4 xOff = 0 yOff = 0 dataOffset[0] 0"
            ]
          },
          "metadata": {},
          "execution_count": 192
        }
      ]
    }
  ]
}