Significant_Features.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy of Project1_starter.ipynb",
      "provenance": [],
      "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/cs167/project-1-niclacey/blob/main/FINAL_Project1_Nicole%20Lacey.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aG0_H2OSeH7w"
      },
      "source": [
        "**Name:** Nicole Lacey"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F9AOhxuO9Vt7"
      },
      "source": [
        "# Project \\#1 Starter Code: \n",
        "Your first step is to decide which dataset you would like to work with for this project. You can find the data on Blackboard. \n",
        "\n",
        "The options include: \n",
        "- [Ramen Ratings](https://www.kaggle.com/residentmario/ramen-ratings)\n",
        "- [Chocolate Ratings](https://www.kaggle.com/rtatman/chocolate-bar-ratings)\n",
        "- [World Happiness Data](https://www.kaggle.com/unsdsn/world-happiness)\n",
        "- [Heart Failure Prediction](https://www.kaggle.com/fedesoriano/heart-failure-prediction)\n",
        "\n",
        "Let's start by taking a look at these different datasets. (you can get rid of this once you decide which one you want to use). "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "5N97IBe79Mfp",
        "outputId": "b723ab2e-4d48-4773-a334-614f053a1375"
      },
      "source": [
        "from google.colab import drive\n",
        "import pandas as pd\n",
        "drive.mount('/content/drive')\n",
        "hap =pd.read_csv('/content/drive/MyDrive/Datasets/happiness_data.csv')\n",
        "hap.head()"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>Overall rank</th>\n",
              "      <th>Year</th>\n",
              "      <th>Country or region</th>\n",
              "      <th>Score</th>\n",
              "      <th>GDP per capita</th>\n",
              "      <th>Social support and Family</th>\n",
              "      <th>Healthy life expectancy</th>\n",
              "      <th>Freedom to make life choices</th>\n",
              "      <th>Generosity</th>\n",
              "      <th>Perceptions of corruption</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>2018</td>\n",
              "      <td>Finland</td>\n",
              "      <td>7.632</td>\n",
              "      <td>1.305</td>\n",
              "      <td>1.592</td>\n",
              "      <td>0.874</td>\n",
              "      <td>0.681</td>\n",
              "      <td>0.202</td>\n",
              "      <td>0.393</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>2018</td>\n",
              "      <td>Norway</td>\n",
              "      <td>7.594</td>\n",
              "      <td>1.456</td>\n",
              "      <td>1.582</td>\n",
              "      <td>0.861</td>\n",
              "      <td>0.686</td>\n",
              "      <td>0.286</td>\n",
              "      <td>0.340</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>2018</td>\n",
              "      <td>Denmark</td>\n",
              "      <td>7.555</td>\n",
              "      <td>1.351</td>\n",
              "      <td>1.590</td>\n",
              "      <td>0.868</td>\n",
              "      <td>0.683</td>\n",
              "      <td>0.284</td>\n",
              "      <td>0.408</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>2018</td>\n",
              "      <td>Iceland</td>\n",
              "      <td>7.495</td>\n",
              "      <td>1.343</td>\n",
              "      <td>1.644</td>\n",
              "      <td>0.914</td>\n",
              "      <td>0.677</td>\n",
              "      <td>0.353</td>\n",
              "      <td>0.138</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>2018</td>\n",
              "      <td>Switzerland</td>\n",
              "      <td>7.487</td>\n",
              "      <td>1.420</td>\n",
              "      <td>1.549</td>\n",
              "      <td>0.927</td>\n",
              "      <td>0.660</td>\n",
              "      <td>0.256</td>\n",
              "      <td>0.357</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Overall rank  Year  ... Generosity  Perceptions of corruption\n",
              "0             1  2018  ...      0.202                      0.393\n",
              "1             2  2018  ...      0.286                      0.340\n",
              "2             3  2018  ...      0.284                      0.408\n",
              "3             4  2018  ...      0.353                      0.138\n",
              "4             5  2018  ...      0.256                      0.357\n",
              "\n",
              "[5 rows x 10 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 1
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7SfueFgsNdNt"
      },
      "source": [
        "# Problem: \n",
        "Which dataset are you going to use? What questions are you going to ask about this dataset? Make sure that you can answer the question using ML models. What are your predictor variables and target variable?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6IWx7uDqYFbc"
      },
      "source": [
        "I am going to use the **happiness** dataset. My target variable is **Score**, and my predictors are everything but Score and Country. I want to know if there is a significant feature in this, or if all of the predictors contribute equally. I will be able to see this based on how well a **tree performs vs. a knn algorithm**. I also want to know how accurate my model is, so I will use **MAE** and **MSE** to see how well it does. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yM_lNmrrNw6S"
      },
      "source": [
        "# Data Preparation\n",
        "Explain your data prep. What did you do to get your data in the shape for your experiments? (e.g. creation of dummy variables, filling in missing values, etc). "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OlBFlXGHbA7p"
      },
      "source": [
        "First, I looked at which columns have missing data. Only one has missing data, so I am going to **create dummy variables** where needed and then put the mean for missing values in **perceptions of corruption**. \n",
        "Note: It turns out no dummy variables are needed. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GyNMqAKCdIbr",
        "outputId": "f0d78359-cde7-470c-f0b1-37dadcf1fb2e"
      },
      "source": [
        "print(hap.isna().any())"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Overall rank                    False\n",
            "Year                            False\n",
            "Country or region               False\n",
            "Score                           False\n",
            "GDP per capita                  False\n",
            "Social support and Family       False\n",
            "Healthy life expectancy         False\n",
            "Freedom to make life choices    False\n",
            "Generosity                      False\n",
            "Perceptions of corruption        True\n",
            "dtype: bool\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        },
        "id": "LLn1FBhicSET",
        "outputId": "cdecc2d9-3c7e-41ba-c360-9bcf742426b7"
      },
      "source": [
        "hap.head()"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Overall rank</th>\n",
              "      <th>Year</th>\n",
              "      <th>Country or region</th>\n",
              "      <th>Score</th>\n",
              "      <th>GDP per capita</th>\n",
              "      <th>Social support and Family</th>\n",
              "      <th>Healthy life expectancy</th>\n",
              "      <th>Freedom to make life choices</th>\n",
              "      <th>Generosity</th>\n",
              "      <th>Perceptions of corruption</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>2018</td>\n",
              "      <td>Finland</td>\n",
              "      <td>7.632</td>\n",
              "      <td>1.305</td>\n",
              "      <td>1.592</td>\n",
              "      <td>0.874</td>\n",
              "      <td>0.681</td>\n",
              "      <td>0.202</td>\n",
              "      <td>0.393</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>2018</td>\n",
              "      <td>Norway</td>\n",
              "      <td>7.594</td>\n",
              "      <td>1.456</td>\n",
              "      <td>1.582</td>\n",
              "      <td>0.861</td>\n",
              "      <td>0.686</td>\n",
              "      <td>0.286</td>\n",
              "      <td>0.340</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>2018</td>\n",
              "      <td>Denmark</td>\n",
              "      <td>7.555</td>\n",
              "      <td>1.351</td>\n",
              "      <td>1.590</td>\n",
              "      <td>0.868</td>\n",
              "      <td>0.683</td>\n",
              "      <td>0.284</td>\n",
              "      <td>0.408</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>2018</td>\n",
              "      <td>Iceland</td>\n",
              "      <td>7.495</td>\n",
              "      <td>1.343</td>\n",
              "      <td>1.644</td>\n",
              "      <td>0.914</td>\n",
              "      <td>0.677</td>\n",
              "      <td>0.353</td>\n",
              "      <td>0.138</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>2018</td>\n",
              "      <td>Switzerland</td>\n",
              "      <td>7.487</td>\n",
              "      <td>1.420</td>\n",
              "      <td>1.549</td>\n",
              "      <td>0.927</td>\n",
              "      <td>0.660</td>\n",
              "      <td>0.256</td>\n",
              "      <td>0.357</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Overall rank  Year  ... Generosity  Perceptions of corruption\n",
              "0             1  2018  ...      0.202                      0.393\n",
              "1             2  2018  ...      0.286                      0.340\n",
              "2             3  2018  ...      0.284                      0.408\n",
              "3             4  2018  ...      0.353                      0.138\n",
              "4             5  2018  ...      0.256                      0.357\n",
              "\n",
              "[5 rows x 10 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "l5nj8HyyJixp"
      },
      "source": [
        "I am realizing that there is one country for every score, so I am just not going to use that column at all. It's not necessary to guess the score. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QF9XAeaDd_d1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "outputId": "0102c07f-b4ed-42a6-a8e7-c32a51294ded"
      },
      "source": [
        "#Filling in missing values with the average\n",
        "corrupt_mean = hap['Perceptions of corruption'].mean()\n",
        "hap['Perceptions of corruption'].fillna(corrupt_mean, inplace=True)\n",
        "print(\"after: \", hap['Perceptions of corruption'].isna().any())\n",
        "hap.head(7)\n"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "after:  False\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>Overall rank</th>\n",
              "      <th>Year</th>\n",
              "      <th>Country or region</th>\n",
              "      <th>Score</th>\n",
              "      <th>GDP per capita</th>\n",
              "      <th>Social support and Family</th>\n",
              "      <th>Healthy life expectancy</th>\n",
              "      <th>Freedom to make life choices</th>\n",
              "      <th>Generosity</th>\n",
              "      <th>Perceptions of corruption</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>2018</td>\n",
              "      <td>Finland</td>\n",
              "      <td>7.632</td>\n",
              "      <td>1.305</td>\n",
              "      <td>1.592</td>\n",
              "      <td>0.874</td>\n",
              "      <td>0.681</td>\n",
              "      <td>0.202</td>\n",
              "      <td>0.393</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>2018</td>\n",
              "      <td>Norway</td>\n",
              "      <td>7.594</td>\n",
              "      <td>1.456</td>\n",
              "      <td>1.582</td>\n",
              "      <td>0.861</td>\n",
              "      <td>0.686</td>\n",
              "      <td>0.286</td>\n",
              "      <td>0.340</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>2018</td>\n",
              "      <td>Denmark</td>\n",
              "      <td>7.555</td>\n",
              "      <td>1.351</td>\n",
              "      <td>1.590</td>\n",
              "      <td>0.868</td>\n",
              "      <td>0.683</td>\n",
              "      <td>0.284</td>\n",
              "      <td>0.408</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>2018</td>\n",
              "      <td>Iceland</td>\n",
              "      <td>7.495</td>\n",
              "      <td>1.343</td>\n",
              "      <td>1.644</td>\n",
              "      <td>0.914</td>\n",
              "      <td>0.677</td>\n",
              "      <td>0.353</td>\n",
              "      <td>0.138</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>2018</td>\n",
              "      <td>Switzerland</td>\n",
              "      <td>7.487</td>\n",
              "      <td>1.420</td>\n",
              "      <td>1.549</td>\n",
              "      <td>0.927</td>\n",
              "      <td>0.660</td>\n",
              "      <td>0.256</td>\n",
              "      <td>0.357</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>6</td>\n",
              "      <td>2018</td>\n",
              "      <td>Netherlands</td>\n",
              "      <td>7.441</td>\n",
              "      <td>1.361</td>\n",
              "      <td>1.488</td>\n",
              "      <td>0.878</td>\n",
              "      <td>0.638</td>\n",
              "      <td>0.333</td>\n",
              "      <td>0.295</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>7</td>\n",
              "      <td>2018</td>\n",
              "      <td>Canada</td>\n",
              "      <td>7.328</td>\n",
              "      <td>1.330</td>\n",
              "      <td>1.532</td>\n",
              "      <td>0.896</td>\n",
              "      <td>0.653</td>\n",
              "      <td>0.321</td>\n",
              "      <td>0.291</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Overall rank  Year  ... Generosity  Perceptions of corruption\n",
              "0             1  2018  ...      0.202                      0.393\n",
              "1             2  2018  ...      0.286                      0.340\n",
              "2             3  2018  ...      0.284                      0.408\n",
              "3             4  2018  ...      0.353                      0.138\n",
              "4             5  2018  ...      0.256                      0.357\n",
              "5             6  2018  ...      0.333                      0.295\n",
              "6             7  2018  ...      0.321                      0.291\n",
              "\n",
              "[7 rows x 10 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y3ha10SpN27E"
      },
      "source": [
        "# Metrics\n",
        "Identify which metrics you will be using to test your model and say why they are appropriate."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kv8aq-tQhC-w"
      },
      "source": [
        "I will be using metrics appropriate for a regression, so **MSE**, **MAE**, and **SD**. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "67v8fsFBN7TR"
      },
      "source": [
        "# Baseline Performance\n",
        "Identify a baseline for your metrics - what would you expect to happen with this particular data if you always guessed the average or if you guessed randomly? What did some of the unsuccessful machine learning algorithms yield?\n",
        "\n",
        "Some helpful links:\n",
        "- [Sklearn Dummy Regressor](https://scikit-learn.org/stable/modules/generated/sklearn.dummy.DummyRegressor.html)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mGE56w6bN5_N",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7ec12f3b-39d2-46a4-ba89-ee4135da9d61"
      },
      "source": [
        "# build and test a baseline model here\n",
        "hap['Score'].mean()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "5.379017902998729"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5-2vpKzaiXyP"
      },
      "source": [
        "I would guess the average is 5, because it is a scale out of 10. The actual average is 5.379. So any model that can find a closer value will be better than the mean. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bot75U5WOLLT"
      },
      "source": [
        "# Model planning and Execution\n",
        "Identify which learning algorithms you have tried and which important parameters you’ve tuned for each one.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_8UhaRfbjkhW"
      },
      "source": [
        "I am going to be trying a knn, sklearn knn, sknlearn weighted knn, tree, random tree, and then all of that on normalized data as well.  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oTqOhnnGOQOZ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6c2c8f3b-f7e3-43b4-9541-55bd5b1af2fc"
      },
      "source": [
        "# Execute your plan here (build and test your models to answer your question)\n",
        "\n",
        "#first is knn regression \n",
        "\n",
        "shuffled_data = hap.sample(frac= 1, random_state = 41)\n",
        "\n",
        "def knnregression(example, train_data, k):\n",
        "  train_datacopy = train_data.copy()\n",
        "  train_datacopy['distance_to_new'] = numpy.sqrt(\n",
        "      (example['GDP per capita']- train_datacopy['GDP per capita'])**2\n",
        "      +(example['Social support and Family'] - train_datacopy['Social support and Family'])**2\n",
        "      +(example['Generosity'] - train_datacopy['Generosity'])**2\n",
        "      +(example['Freedom to make life choices'] - train_datacopy['Freedom to make life choices'])**2\n",
        "      +(example['Perceptions of corruption'] - train_datacopy['Perceptions of corruption'])**2\n",
        "      +(example['Healthy life expectancy'] - train_datacopy['Healthy life expectancy'])**2)\n",
        "  sorted_data = train_datacopy.sort_values(['distance_to_new'])\n",
        "  return sorted_data.iloc[0:k]['Score'].mean()\n",
        "\n",
        "def MSE(actual,predicted):\n",
        "  squared = []\n",
        "  csquared = (actual - predicted)**2\n",
        "  squared.append(csquared)\n",
        "  MSE = numpy.sum(squared)/(test_data['Score'].count())\n",
        "  return MSE\n",
        "\n",
        "def all_knn_mse(test_data,train_data,k):\n",
        "  return test_data.apply(knnregression,args=(train_data,k),axis=1)\n",
        "\n",
        "test_data = shuffled_data.iloc[0:500]\n",
        "train_data = shuffled_data.iloc[500:]\n",
        "\n",
        "test_data.shape\n",
        "train_data.shape"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(282, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 308
        },
        "id": "PU2IOyxSm4eH",
        "outputId": "55937791-80e0-4eba-86e7-0233b10071a9"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy\n",
        "%matplotlib inline\n",
        "\n",
        "k_vals = [1,3,5,9,15,21,31,51,101,129, 500]\n",
        "knn_mse = []\n",
        "\n",
        "for k in k_vals:\n",
        "    predictions = all_knn_mse(test_data,train_data,k)\n",
        "    current_mse = MSE(test_data['Score'],predictions)\n",
        "    knn_mse.append(current_mse)\n",
        "\n",
        "plt.suptitle('Score',fontsize=18)\n",
        "plt.xlabel('k')\n",
        "plt.ylabel('MSE')\n",
        "plt.plot(k_vals,knn_mse,'ro-',label='k-NN')\n",
        "plt.legend(loc='lower left', shadow=True)\n",
        "plt.axis([0,130,0,1])\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEjCAYAAAA/ugbCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de5wU5Z3v8c9vhosOt1EZQS7DACECohEd1KhZJWrEqJBNzBHFzarE2XNWVz3JZqPhbBLNi7wS3bPGGE0OJor6GvW4npigkhjW6ElyAoQhKgiIAgqMSsQLoI5yyfzOH0+109OXudf0dNf3/Xr1q7urqrufoobnW/U8VU+ZuyMiIslVVugCiIhIYSkIREQSTkEgIpJwCgIRkYRTEIiIJJyCQEQk4RQEIiIJpyCQkmVmE8xskZm9YGZNZvaOmW0ws3vMbGahyyfSV/QrdAFE4mBmtcD/BfYD9wLrgIOBScBngHeBpwpWQJE+REEgpepbQAVwrLs/lznTzEb2doHMbIi7v9vbvyvSHjUNSamaBLyVKwQA3H1H+nszm2lmj5vZW2b2oZltMbOfmdnwtGX6mdnXzWx9tMxbZvaImR2d8V01ZuZm9m0zu9DMVpvZB8BtacucaWa/MbNd0XetMbP/2rP/BCIdoyCQUrUZOMzMPt/egmb2D8CTwDHAj4F/AuqB44ExaYvWA98DGoGvAT8BZgLLzWx6jq/+XPR9vwauBn4V/V4d8BtgMLAQ+EpU3h+b2c2dXVGR7jINOielyMw+Segj6A+8BPwBWAU87e4b0pYbQ6iENwMnu/uujO8pc/dmMzuLUHk/BMz16D+OmX0CWA0sd/dPRdNqgJeBA8AxGb93RDTv5+5+ccZv3QpcBUxy9y099E8h0i4dEUhJcvflhD36e4BhwGXAHcB6M/udmU2IFv0iMAC4ITMEou9pjl7+bfS80NP2nqKmp0eBU82sKuPjj6eHQOQCYCDwMzMbnv6IvqcMOLNray3SNeoslpLl7muBSwHMbBxwGvBl4FPAL83seEJfAsAz7XzdeKAZyKzYIZyR9LlomZ1p01/MseyU6Pk/2/itEe2URaRHKQgkEdx9K3Cvmd0H/B44BTgh5p9tyjHNoucvAa/n+ZyahaRXKQgkUdzdzWwlIQhG07LXfiy59+BTthCabaYAazLmTY2eX+5AEV6Knt9097aOCkR6jfoIpCSZ2VlmlrWjY2YHEy4oA1gPPAzsA75lZkNzLJ/ag/9F9Hx92jTMbBowG/iDu+/M/HwODwF7gRuismT+3jAzG9iB7xHpMToikFJ1C+H00SXAWkIzzVjgYuDjwL1RHwJmdi1wO7DWzO4FthKOFuYAlwPPuvsyM3sImAscYmaPASOBK4EPCaeHtsvdG83svwE/BTZETVVbgSrgaEJfw1TglW7/C4h0kIJAStVXCBX5qcAXgEpgN6FZ5/vA4tSC7v5jM9tMuDbgasJZPa8Rri3Ynvad84A/Ezqg/yfwPuEU1X9NhUpHuPvdZvYi8M/AP0RlexPYCPwrsKONj4v0OF1HICKScOojEBFJOAWBiEjCKQhERBJOQSAiknAKAhGRhFMQiIgknIJARCThFAQiIgmnIBARSTgFgYhIwsUWBGZ2l5m9YWbP55lvZvZDM9sU3bj7uLjKIiIi+cV5RLAYmNXG/HMId4eaBNQRbvItIiK9LLYgcPffAW+3scgcwlDA7u4rgMroxt4iItKLCjkM9WhaD/HbGE3Lun2fmdURjhoYNGjQ8ZMnT+6VAoqIlIrVq1e/6e5VueYVxf0I3H0RsAigtrbWGxoaClwiEZHiYmZb880r5FlDrxLuGJUyJpomIiK9qJBBsAT4UnT20EnAbnfPahYSEZF4xdY0ZGYPAKcDw82sEfgW0B/A3X8CLAU+C2wi3E/2srjKIiIi+cUWBO5+UTvznXDjbxERKSBdWSwiknAKAhGRhFMQiIgknIJARCThFAQiIgmnIBARSTgFgYhIwikIREQSTkEgIpJwCgIRkYRTEIiIJJyCQEQk4RQEIiIJpyAQEUk4BYGISMIpCEREEk5BICKScAoCEZGEUxCIiCScgkBEJOEUBCIiCacgEBFJOAWBiEjCKQhERBJOQSAiknAKAhGRhFMQiIgknIJARCThFAQiIgmnIBARSTgFgYhIwikIREQSTkEgIpJwCgIRkYSLNQjMbJaZbTSzTWZ2XY751Wb2lJk9Y2ZrzOyzcZZHRESyxRYEZlYO3A6cA0wFLjKzqRmL/Q/gIXefDswF7oirPCIiklucRwQnAJvcfYu77wMeBOZkLOPA0Oj1MOC1GMsjIiI59Ivxu0cD29PeNwInZizzbeA3ZvZPwCDgzBjLIyIiORS6s/giYLG7jwE+C9xnZlllMrM6M2sws4adO3f2eiFFREpZnEHwKjA27f2YaFq6+cBDAO6+HDgIGJ75Re6+yN1r3b22qqoqpuKKiCRTnEGwCphkZuPNbAChM3hJxjLbgDMAzGwKIQi0yy8i0otiCwJ3PwBcBTwBbCCcHbTOzG40s9nRYl8FrjCz54AHgEvd3eMqk4iIZIuzsxh3XwoszZj2zbTX64FT4iyDiIi0rdCdxSIiUmAKAhGRhFMQiIgknIJARCThFAQiIgmnIBARSTgFgYhIwikIREQSTkEgIpJwCgIRkYRTEIiIJJyCQEQk4RQEIiIJpyAQEUk4BYGISMIpCEREEk5BICKScAoCEZGEUxCIiCScgkBEJOEUBCIiCacgEBFJOAWBiEjCKQhERBJOQSAiknAKAhGRhFMQiIgknIJARCThFAQiIgmnIBARSTgFgYhIwikIREQSTkEgIpJwCgIRkYRTEIiIJFysQWBms8xso5ltMrPr8izzX8xsvZmtM7P74yyPiIhk6xfXF5tZOXA7cBbQCKwysyXuvj5tmUnA9cAp7v6OmR0eV3lERCS3OI8ITgA2ufsWd98HPAjMyVjmCuB2d38HwN3fiLE8IiKSQ5xBMBrYnva+MZqW7uPAx83s/5nZCjObleuLzKzOzBrMrGHnzp0xFVdEJJkK3VncD5gEnA5cBNxpZpWZC7n7InevdffaqqqqXi6iiEhpizMIXgXGpr0fE01L1wgscff97v4y8CIhGEREpJfEGQSrgElmNt7MBgBzgSUZy/yCcDSAmQ0nNBVtibFMIiKSIbYgcPcDwFXAE8AG4CF3X2dmN5rZ7GixJ4C3zGw98BTwNXd/K64yiYhINnP3QpehU2pra72hoaHQxRARKSpmttrda3PNK3RnsYiIFJiCQEQk4RQEIiIJpyAQEUk4BYGISMIpCEREEk5BICKScAoCEZGEazMIzOyStNenZMy7Kq5CiYhI72nviOAraa9vy5h3eQ+XRURECqC9ILA8r3O9FxGRItReEHie17nei4hIX1BfDzU1UFYWnuvr21y8vXsWTzazNYS9/4nRa6L3E7pbVhER6WH19VBXB01N4f3WrXDFFQyHQ/N9pL0gmNKDxRMRkbhdf31LCKR88AGjsm8V/JE2g8Ddt6a/N7PDgL8Btrn76q6XVEREesRrr8Hy5fDHP4bH9u05F+sPA/J9RZtBYGaPAde5+/NmdgTwZ6CB0Ey0yN1/0I3ii4hIZ+zfD2vWhAo/VflvjfbXBw6EGTNg6FDYsyf7o7Av39e211k83t2fj15fBixz9/OBEynk6aOd7AgRESlKb74Jjz4K3/gGnH46VFZCbS1cfTX87ndwwglwyy2wcmWo/H//e7jjDqioaP09FRW8ln3P+I+010ewP+31GcCdAO7+rpk1d23Nuuntt7M7Qurqwut58wpSJBGRbmtuhvXrW5p4li+HF18M8/r1g+nT4Yor4JOfhJNPhrFjc39Pqh5csAC2bYPqali4kDcvueTtfD/d5q0qzexR4DdAI3AX4Qhhl5kdDDS4+1FdWd/uqB040Bv25TjCGTcOXnmlt4sjItI1u3eHPflUE8+KFS1NOlVVobJPVfrHH5+9l99Jbd2qsr0jgvnAjcCZwIXuviuafhJwd7dK1VW5QgBC8omI9EXu8NJLrTt1160L08vKYNo0uPjilsp/4kSw3rtmt/huXq8jAhHp65qaYNWqliae5ctDez/AsGGhsk/t7Z9wQujgjVmXjwjMbElb8919dncK1iWjR8Nf/tL6PNmKCli4sNeLIiKCe2iRSD+T59ln4a9/DfMnT4bzzw+V/sknh/dlfWvg5/aahj4JbAceAFbSF8YXOvRQ+M534B//MbSnVVfDd7+rjmIR6R1798Kf/9xS6S9fHs7lBxg0KOzhX3ddqPRPOinUWX1ce0EwEjgLuAi4GHgceMDd18VdsDbNmwcffghf/jL89rehPU1EJA6vv9660m9oaOmrHD8eZs5saeY5+uhwhk+Rae/K4r8CvwZ+bWYDCYHwtJnd4O4/6o0C5jUlGv1iwwYFgYj0jAMHwgVb6Z26qb7HgQNbzuFPdeqOHFnQ4vaUdqMrCoBzCSFQA/wQeCTeYnVAKgheeAHOO6+wZRGR4vTWW+G0zdTe/sqVLf2Po0aFCv/qq0OlP316CIMS1F5n8b3ANGApcEPaVcaFd8ghMGJEOCIQEWlPc3OoL9I7dTduDPPKy0NFP39+S6fu2LG9egpnIbV3RHAJ8D5wDXC1tfyjGODuHv85T22ZMkVBIJJE9fVZV85mnTCyZw/86U8tTTwrVoSLuACGDw97+ZdeGir92tpuX7BVzNrrI+hb5zhlmjwZHnwwnL6VkOQWSbxc4+3X1YXTyocPb9njX7u2pW6YNg3mzm3p1P3Yx1RnpCm+7u10U6bArl3hD6BEOm1EpB25xttvaoKvfjW8Hjo0VPif/3zLBVvDhvV+OYtI8QcBhA5jBYFIaXAPnbhbtsDmzdnPjY35P7t2LUyd2ucu2OrrSiMINmwIQ7SKSHE4cCC07+er7DPH0z/iiHCa+Kc/Db/8ZUtbf7px40ITkHRacQfB6NEweLA6jEX6onffzV3Jb9kS2vUPHGhZdsCAcHHWxIlwyinhecKE8Dx+fOuO3Mw+AtAwM91U3EFgFjqMFQQiva+5OVx1m6uy37y5ZZC1lMMOC5X7jBlw4YWtK/tRo8IpnB2RZ7x9DTPTdbEGgZnNAm4FyoGfuvv38iz3BeBhYIa7N3TqR6ZMCcNMiEjP+/BDePnl3JX9yy+H+SllZaF5ZsKE0FGbquQnTAiPysqeK9e8ear4e1BsQWBm5cDthLGKGoFVZrbE3ddnLDeEcJ3Cyi790JQpcN994TB0yJBullokYVIds6m9+MzK/tWMuxsOHhwq9cmT4dxzW1f248ZB//6FWQ/pljiPCE4ANrn7FgAzexCYA6zPWO47wPeBr3XpV9LPHJoxo4tFFSlh+/e33TH77rutlx81KlTuZ57ZuvlmwoRw5yydf19y4gyC0YQhrFMaCTe9/4iZHQeMdffHzaxrQTB5cnjesEFBIMm1Z0/bHbOpsfEhjJeT6pj91KdaV/Y1NYm+wjapCtZZbGZlwL8Dl3Zg2TqgDqC6urr1zIkTw7Cv6jCWUtbcHMa8z1fZ5+qYnTgRTjwRLroou2NW59lLmjiD4FVgbNr7MdG0lCGEAe2ejsYwGgksMbPZmR3G7r4IWARQW1vb+t6a/fvDpEkKAil+H3zQdsfs3r0ty5aXh7NlJk6EL3whu2NWV9JKJ8QZBKuASWY2nhAAcwk3twHA3XcDw1Pvzexp4J87fdYQhH6CdYW9V45Iu9zDnnu+jtnUXa5SBg8OlfvUqeFWh+mVfXW1Omalx8QWBO5+wMyuAp4gnD56l7uvM7MbgQZ3b/N+yJ0yeXK42nDfvnBhikihpDpm81X2773XevnRo0PF/pnPZHfMDh+ujlnpFbH2Ebj7UsK9DNKnfTPPsqd3+YemTAmdYZs2hb0nkTjt3p2/rX7btuyO2VRzzWmnZXfMHnxwwVZDJKW4ryxOSR9zSEEg3dXcHM6fz1fZv/VW6+WHDw8V+0knhYuc0iv7I45Qx6z0eaURBOmnkIp0RFNT2x2zqZuTQ+iYHTcuVOwXXBCe0ztmhxb2/kwi3VUaQTBoUOg8e+GFQpdEekNH7k7lDm+8kf8iqtdfb738kCGhcp82DWbPbl3ZV1eHU5RFSlTp/HVr8LlkyHV3qvnzYdmy0ESTquy3bMnumB0zJlTss2a1NN2kKvvDDlPHrCRW6QTBlClw552hfVdtssXPHd55J+z1px7bt8OPfpR9d6q9e+Gee+Cgg1qaa2bObF3Z19SE+SKSpXSCYM+eUEH066dhaYvB3r3hTlOZFX36+/ffb/2ZAQNat92nMwvLaydApNNKIwjq6+H++8Nr95abWYPCoBDcYefO1pV6ZmW/Y0f250aMCCE+ZQqcfXZ4nf6oqgp7+Vu3Zn+2ulohINJFpREECxa0vvwewtHBggUKgjg0NWXvvWdW9Jnbo6KipUI/+ujsSn7MmI413SxcqLtTifSw0giCbds6Nz2JOnKmDYQ+lh07clfyqYo+c4AzszCQWXU1HHccfO5zMHZs64r+0EN7pjNWd6cS6XHm7u0v1YfU1tZ6Q0PGcEQ1NbmbC4YOhUMOUYWR6x6vAwfCxReHC57SK/pXXw3DJKQbMiScR59esadX9KNHa9wbkT7OzFa7e23OeSURBLkqulwqKmDRor4VBh3dU2/Le+/BX/6S//H449lNNSnl5aFZJrOpJr2y10iWIkWvrSAojaahXM0F772XPRRAUxNcc01ovjjyyJbOxZ6ojLsi1znxqU7u885ru3JPf+QLwMMOCx2w+ULALMzr6E3DRaQklcYRQS5lZeHslXyGDQs37aiogF/9qnVl2dkjh84EyQcfhNMmGxvhi1/MDqu2mIWLpkaMaP9x+OEtzTX5ms7GjYNXXun474tI0Sr9I4JcqqtzV35HHAHf/S6sWBEezz2XvUxTE1xxRbhatbIyPA45pOV1+mPZsnCUkXml6x//GMqwfXvrR2ZHaz433QQjR7au3IcP79pQBzrTRkTaULpHBLn6DXLt6bd15FBdDbt2hYvVuqqyMrTBjx2b/fi7v8u+GQnEs6deqOYvEekTknlE0NHTDPMdOaRXxgcOhDDYtSv7MX9+7t83C+PWDxmSv4w33dR7e+rz5qniF5GcSveIoKM6euSQT3fb37WnLiK9oK0jAl2TP29eqPTHjQt78ePGda6jeOHCEBzpOrNXP29eCIzm5vCsEBCRXla6TUOd0Z1mE13pKiJFTkHQE9T+LiJFTE1DIiIJpyAQEUk4BYGISMIpCEREEk5BICKScAoCEZGEUxCIiCScgkBEJOEUBCIiCacgEBFJOAWBiEjCKQhERBJOQSAiknAKAhGRhFMQiIgkXKxBYGazzGyjmW0ys+tyzP+Kma03szVm9qSZjYuzPCIiki22IDCzcuB24BxgKnCRmU3NWOwZoNbdjwEeBm6KqzwiIpJbnEcEJwCb3H2Lu+8DHgTmpC/g7k+5e+qu8SuAMTGWR0REcogzCEYD29PeN0bT8pkP/CrXDDOrM7MGM2vYuXNnDxZRRET6RGexmV0C1AI355rv7ovcvdbda6uqqnq3cCIiJS7Om9e/CoxNez8mmtaKmZ0JLABOc/e9MZZHRERyiPOIYBUwyczGm9kAYC6wJH0BM5sO/C9gtru/EWNZREQkj9iCwN0PAFcBTwAbgIfcfZ2Z3Whms6PFbgYGA/9hZs+a2ZI8XyciIjGJs2kId18KLM2Y9s2012fG+fsiItK+PtFZLCIihaMgEBFJOAWBiEjCKQhERBJOQSAiknAKAhGRhFMQiIgknIJARCThFAQiIgmnIBARSTgFgYhIwikIREQSTkEgIpJwCgIRkYRTEIiIJFys9yPoLfv27WPz5s00NTUVuiixq6ioYOLEiQwYMKDQRRGRElESQbB582YqKys58sgjKSsr3YOc5uZmduzYwXPPPcdBBx3EtGnTMLNCF0tEilxJ1JpNTU2MGDGipEMAoKysjJEjR2JmLFu2jJUrVxa6SCJSAkqm5iz1EEgpKyvDzDj00ENZs2ZNoYsjIiUgGbVnL3jllVeYNm1am8tceumljB49mr179wLw5ptvUlNT89HnzYzbbrvto+WvuuoqFi9enPO7ysvL2b9/f4+UXUSSLZlBUF8PNTVQVhae6+t77afLy8u56667cs47/PDDufXWW9m3b1+vlUdEJHlBUF8PdXWwdSu4h+e6uh4Ngy1btjB9+nRWrVqVNe/aa6/llltu4cCBA1nzqqqqOOOMM7jnnnt6rCwiIu0pibOGWrn2Wnj22fzzV6yAqGnmI01NMH8+3Hln7s8ceyz84Acd+vmNGzcyd+5cFi9ezCc+8Yms+dXV1Zx66qncd999nH/++Vnzv/71r3POOedw+eWXd+j3RES6K3lHBJkh0N70Tti5cydz5syhvr4+ZwikXH/99dx88800NzdnzZswYQInnngi999/f7fLIyLSEaV3RNDenntNTWgOyjRuHDz9dLd+etiwYVRXV/OHP/yBqVOnctlll/HMM88watQoli5d+tFykyZN4thjj+Whhx7K+T3f+MY3uOCCCzjttNO6VR4RkY4ovSBoz8KFoU8g/SrkioowvZsGDBjAI488wtlnn83gwYO5++678y67YMECzj333JzzJk+ezNSpU3n00UeZMWNGt8slItKW5DUNzZsHixaFIwCz8LxoUZjeAwYNGsRjjz3GLbfcwpIlS/Iud9RRR3Hcccflnb9gwQIaGxt7pEwiIm0xdy90GTqltrbWGxoaWk1bvXo1xx9/fIFK1PtWr17N+vXr2bNnD1deeWWhiyMiRcDMVrt7ba55yTsiEBGRVhQEIiIJpyAQEUm4kgmCXOfkl6KkrKeI9J6SCIKKigp27NhR8pVk6n4EqcHmdC8CEekJJXEdwcSJE3nhhRd47bXXSr5y3L9/P9u2beP9999n6NChhS6OiJSAWIPAzGYBtwLlwE/d/XsZ8wcC9wLHA28BF7r7K539nQEDBnDUUUfx9NNPs3bt2kTcm6B///6cddZZhS6GiJSA2ILAzMqB24GzgEZglZktcff1aYvNB95x94+Z2Vzg+8CFXfm98vJyZs6cyTHHHPPReP+lqqysjMrKSioqKgpdFBEpAXEeEZwAbHL3LQBm9iAwB0gPgjnAt6PXDwM/MjPzLl7lVlZWRlVVVddLLCKSQHG2oYwGtqe9b4ym5VzG3Q8Au4HDYiyTiIhkKIrOYjOrA+qit3vN7PlClqeHDQfeLHQhekgprQuU1vqU0rpAaa1Pb63LuHwz4gyCV4Gxae/HRNNyLdNoZv2AYYRO41bcfRGwCMDMGvKNl1GMSml9SmldoLTWp5TWBUprffrCusTZNLQKmGRm481sADAXyByOcwnw99HrC4DfdrV/QEREuia2IwJ3P2BmVwFPEE4fvcvd15nZjUCDuy8BfgbcZ2abgLcJYSEiIr0o1j4Cd18KLM2Y9s201x8CX+zk1y7qgaL1JaW0PqW0LlBa61NK6wKltT4FX5eiux+BiIj0rNK/BFdERNpUVEFgZrPMbKOZbTKz6wpdns4ws7Fm9pSZrTezdWZ2TTT9UDNbZmYvRc+HFLqsHWVm5Wb2jJk9Fr0fb2Yro+3zv6OTBIqCmVWa2cNm9oKZbTCzTxbrtjGz/x79jT1vZg+Y2UHFtG3M7C4zeyP9NPF828KCH0brtcbM8t//tUDyrM/N0d/aGjN7xMwq0+ZdH63PRjM7uzfKWDRBkDZkxTnAVOAiM5ta2FJ1ygHgq+4+FTgJuDIq/3XAk+4+CXgyel8srgE2pL3/PnCLu38MeIcwhEixuBX4tbtPBj5BWK+i2zZmNhq4Gqh192mEEzVSw7cUy7ZZDMzKmJZvW5wDTIoedcCPe6mMnbGY7PVZBkxz92OAF4HrAaI6YS5wVPSZO6K6L1ZFEwSkDVnh7vuA1JAVRcHdX3f3P0ev3yVUNKMJ63BPtNg9wOcKU8LOMbMxwLnAT6P3BnyaMFQIFNe6DAP+hnAWG+6+z913UaTbhnASyMHRtTkVwOsU0bZx998RziJMl29bzAHu9WAFUGlmR/ROSTsm1/q4+2+i0RQAVhCus4KwPg+6+153fxnYRKj7YlVMQdCRISuKgpnVANOBlcAId389mrUDGFGgYnXWD4B/AVI3gTgM2JX2x11M22c8sBO4O2rq+qmZDaIIt427vwr8G7CNEAC7gdUU77ZJybctSqFeuBz4VfS6IOtTTEFQEsxsMPB/gGvdfU/6vOhiuj5/GpeZnQe84e6rC12WHtIPOA74sbtPB94noxmoiLbNIYS9yvHAKGAQ2c0SRa1YtkVHmNkCQrNxfSHLUUxB0JEhK/o0M+tPCIF6d/95NPkvqUPZ6PmNQpWvE04BZpvZK4Qmuk8T2tgro+YIKK7t0wg0uvvK6P3DhGAoxm1zJvCyu+909/3Azwnbq1i3TUq+bVG09YKZXQqcB8xLG1GhIOtTTEHQkSEr+qyoDf1nwAZ3//e0WenDbPw98MveLltnufv17j7G3WsI2+G37j4PeIowVAgUyboAuPsOYLuZHRlNOoMwXHrRbRtCk9BJZlYR/c2l1qUot02afNtiCfCl6Oyhk4DdaU1IfZaFm3b9CzDb3ZvSZi0B5prZQDMbT+gE/1PsBXL3onkAnyX0sG8GFhS6PJ0s+6mEw9k1wLPR47OEtvUngZeA/wQOLXRZO7lepwOPRa8nRH+0m4D/AAYWunydWI9jgYZo+/wCOKRYtw1wA/AC8DxwHzCwmLYN8AChf2M/4Whtfr5tARjhbMLNwFrC2VIFX4cOrM8mQl9Aqi74SdryC6L12Qic0xtl1JXFIiIJV0xNQyIiEgMFgYhIwikIREQSTkEgIpJwCgIRkYRTEIh0k5nVpI8sKVJsFAQiIgmnIBDpQWY2IRq4bkahyyLSUbHes1gkSaIhKh4ELnX35wpdHpGOUhCI9Iwqwvg3n3f39YUujEhnqGlIpGfsJgz4dmqhCyLSWToiEOkZ+4C/BZ4ws/fc/f5CF0ikoxQEIj3E3d+PbtqzLAqDohkmXZJNo4+KiCSc+ghERBJOQSAiknAKAhGRhFMQiIgknIJARCThFAQiIgmnIBARSTgFgYhIwv1/PTRxABEs4LAAAAAASURBVDWvBh4AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qXtzOt7ZpLEw"
      },
      "source": [
        "import sklearn\n",
        "import pandas\n",
        "from sklearn import tree\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn import metrics\n",
        "from sklearn import neighbors\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "#split the data\n",
        "predictors = ['GDP per capita',\t'Social support and Family',\t'Healthy life expectancy', \n",
        "              'Freedom to make life choices',\t'Generosity',\t'Perceptions of corruption']\n",
        "target = \"Score\"\n",
        "train_data, test_data, train_sln, test_sln = \\\n",
        "        train_test_split(hap[predictors], hap[target], test_size = 0.2, random_state=41)\n"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nzcTzJgxpxGr",
        "outputId": "f64b6f08-5df1-48e6-ae63-7982ef493afc"
      },
      "source": [
        "# make sure it looks okay\n",
        "print('train_data shape: ',train_data.shape)\n",
        "print('test_data shape: ',test_data.shape)\n",
        "print('train_sln shape: ',train_sln.shape)\n",
        "print('test_sln shape: ',test_sln.shape)\n",
        "\n",
        "print(train_data.head())\n",
        "print(test_data.head())\n",
        "print(train_sln.head())\n",
        "print(test_sln.head())"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "train_data shape:  (625, 6)\n",
            "test_data shape:  (157, 6)\n",
            "train_sln shape:  (625,)\n",
            "test_sln shape:  (157,)\n",
            "     GDP per capita  ...  Perceptions of corruption\n",
            "294        0.275000  ...                   0.085000\n",
            "555        1.027870  ...                   0.212860\n",
            "584        0.740360  ...                   0.250280\n",
            "347        1.070622  ...                   0.046669\n",
            "660        1.207400  ...                   0.135860\n",
            "\n",
            "[5 rows x 6 columns]\n",
            "     GDP per capita  ...  Perceptions of corruption\n",
            "333        1.107353  ...                   0.111093\n",
            "603        1.094260  ...                   0.124250\n",
            "143        0.357000  ...                   0.099000\n",
            "700        0.474280  ...                   0.042320\n",
            "20         1.233000  ...                   0.034000\n",
            "\n",
            "[5 rows x 6 columns]\n",
            "294    4.085\n",
            "555    5.155\n",
            "584    4.404\n",
            "347    6.357\n",
            "660    6.302\n",
            "Name: Score, dtype: float64\n",
            "333    6.635\n",
            "603    3.974\n",
            "143    3.692\n",
            "700    5.286\n",
            "20     6.711\n",
            "Name: Score, dtype: float64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zkYb9Y5gqLa3",
        "outputId": "b2a305b1-046d-4092-8e79-2d48bc2d1ad4"
      },
      "source": [
        "# kNN \n",
        "# create our model\n",
        "neigh = neighbors.KNeighborsRegressor(n_neighbors=10)\n",
        "\n",
        "# fit (train) the model to the data\n",
        "neigh.fit(train_data, train_sln)\n",
        "\n",
        "# use the trained model to get predictions from our test_data\n",
        "predictions = neigh.predict(test_data)\n",
        "\n",
        "# use a metric to see how good our predictions are\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.2583534494418564\n",
            "r2:  0.7907441919640464\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "r9-XFWeZqOMl",
        "outputId": "50eb39bd-dced-4ac6-87c0-c98781884012"
      },
      "source": [
        "# w-knn \n",
        "# create our model\n",
        "neigh = neighbors.KNeighborsRegressor(n_neighbors=10, weights = 'distance')\n",
        "\n",
        "# fit (train) the model to the data\n",
        "neigh.fit(train_data, train_sln)\n",
        "\n",
        "# use the trained model to get predictions from our test_data\n",
        "predictions = neigh.predict(test_data)\n",
        "\n",
        "# use a metric to see how good our predictions are\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.24670061968015633\n",
            "r2:  0.8001825111076766\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h1jzL2HOqrXO",
        "outputId": "d20f39b9-208b-403d-b1bc-98653dad701b"
      },
      "source": [
        "decision = tree.DecisionTreeRegressor(criterion='mse', max_leaf_nodes= 8)\n",
        "decision.fit(train_data, train_sln)\n",
        "predictions = decision.predict(test_data)\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.3942506709625031\n",
            "r2:  0.6806729583088371\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 683
        },
        "id": "-kEi-V74qwn0",
        "outputId": "b17f32ee-a671-4134-d7a9-50cd1a9ed1c9"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.figure(figsize=(12,12)) # Makes is to the graph isn't tiny\n",
        "tree.plot_tree(decision);"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x864 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1NRdLi83qKPC",
        "outputId": "5c393cf0-a0fb-435b-f27e-517dd15c12cf"
      },
      "source": [
        "from sklearn.ensemble import RandomForestRegressor\n",
        "forest = RandomForestRegressor(random_state = 0, n_estimators = 100)\n",
        "forest.fit(train_data, train_sln)\n",
        "predictions = forest.predict(test_data)\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.22085768282030896\n",
            "r2:  0.8211142410548185\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "exA1ulkNr2W0"
      },
      "source": [
        "This is all on **normalized** data after here. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tTWzNpy_r1uG",
        "outputId": "ca0159e9-59e4-49e4-80c7-a9db53b100eb"
      },
      "source": [
        "# normalize the data\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "scaler = StandardScaler()\n",
        "scaler.fit(train_data)\n",
        "train_data_normalized = scaler.transform(train_data)\n",
        "test_data_normalized = scaler.transform(test_data)\n",
        "\n",
        "# run a knn\n",
        "# using the nonweighted with small k \n",
        "neigh = neighbors.KNeighborsRegressor(n_neighbors=10)\n",
        "\n",
        "# fit (train) the model to the data\n",
        "neigh.fit(train_data_normalized, train_sln)\n",
        "\n",
        "# use the trained model to get predictions from our test_data\n",
        "predictions = neigh.predict(test_data_normalized)\n",
        "\n",
        "# use a metric to see how good our predictions are\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.2570793186662046\n",
            "r2:  0.7917761861780912\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cd7x5K6csdh5",
        "outputId": "2dc5949e-600f-4beb-e33e-69da2ea0446d"
      },
      "source": [
        "neigh = neighbors.KNeighborsRegressor(n_neighbors=10, weights = 'distance')\n",
        "\n",
        "# fit (train) the model to the data\n",
        "neigh.fit(train_data_normalized, train_sln)\n",
        "\n",
        "# use the trained model to get predictions from our test_data\n",
        "predictions = neigh.predict(test_data_normalized)\n",
        "\n",
        "# use a metric to see how good our predictions are\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.24917127767644975\n",
            "r2:  0.7981813784093836\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YLr12gFWsx3t",
        "outputId": "f614f43a-fbee-4928-d212-ddeb41e296bf"
      },
      "source": [
        "decision = tree.DecisionTreeRegressor(criterion='mse', max_leaf_nodes= 8)\n",
        "decision.fit(train_data_normalized, train_sln)\n",
        "predictions = decision.predict(test_data_normalized)\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.3942506709625031\n",
            "r2:  0.680672958308837\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 683
        },
        "id": "CKt9hH8Bs8hP",
        "outputId": "3d1f3a12-f1e1-41de-e272-9d4eb629f595"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.figure(figsize=(12,12)) # Makes is to the graph isn't tiny\n",
        "tree.plot_tree(decision);"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x864 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ddJ_LGPH0FAU",
        "outputId": "c5c47716-7827-44ca-8695-6cb06b109592"
      },
      "source": [
        "forest = RandomForestRegressor(random_state = 0, n_estimators = 100, max_features = 3)\n",
        "forest.fit(train_data_normalized, train_sln)\n",
        "predictions = forest.predict(test_data_normalized)\n",
        "print('MSE: ', metrics.mean_squared_error(test_sln, predictions))\n",
        "print('r2: ', metrics.r2_score(test_sln, predictions))"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MSE:  0.22775472068951289\n",
            "r2:  0.8155279203167256\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kMaqpNm4O9Qy"
      },
      "source": [
        "# Results\n",
        "After you conduct your learning experiment, summarize the results you got. Include visualizations (i.e. graphs) where appropriate.  How did you find the best k, n, m or other variables that you needed to tune. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-hoeqVRIsIP7"
      },
      "source": [
        "Type     -            MSE   -   R^2 \\\n",
        "Manual Knn    -       .3-\\\n",
        "Knn            -      .25    -  .79\\\n",
        "W-Knn - .2467 - .8 \\\n",
        "Decision Tree - .39 - .68 \\\n",
        "**Forest - .2208 - .82** \\\n",
        "Normalized Knn - .257 - .791 \\\n",
        "Normalized W-Knn - .249 - .798 \\\n",
        "Normalized Decision Tree - .394 - .68\\\n",
        "Normalized Forest - .225 - .817    "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 325
        },
        "id": "DJW5f-tt1L4j",
        "outputId": "ad709313-6293-4b27-db26-d0f3c4e23af4"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "def tune_number_of_trees(n_estimator_values):\n",
        "    randomforest_MSE = []\n",
        "    for n in n_estimator_values:\n",
        "        curr_randomforest = RandomForestRegressor(n_estimators=n, random_state=41)\n",
        "        curr_randomforest.fit(train_data,train_sln)\n",
        "        curr_predictions = curr_randomforest.predict(test_data)\n",
        "        curr_MSE = metrics.mean_squared_error(test_sln,curr_predictions)\n",
        "        randomforest_MSE.append(curr_MSE)\n",
        "\n",
        "    plt.suptitle('Random Forest MSE vs. Number of Trees', fontsize=20)\n",
        "    plt.xlabel('Number of trees')\n",
        "    plt.ylabel('MSE')\n",
        "    plt.plot(n_estimator_values,randomforest_MSE,'ro-', label = \"MSE\")\n",
        "    plt.axis([0,n_estimator_values[-1]+1,0,1.0])\n",
        "    plt.legend()\n",
        "    plt.show()\n",
        "\n",
        "    print(min(randomforest_MSE))\n",
        "    \n",
        "tune_number_of_trees(range(1,150))"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.22635385202835476\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Vfefn_8X5JQN"
      },
      "source": [
        "MSEs lowest is at 18 so I will use 18 trees earlier in the project. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 325
        },
        "id": "wHA1I6cAOj6P",
        "outputId": "9fc8f271-f0d5-4afc-f5aa-16cacc7de25f"
      },
      "source": [
        "def tune_max_features(max_features_values):\n",
        "    randomforest_MSEs = []\n",
        "\n",
        "    for m in max_features_values:\n",
        "        currrf = RandomForestRegressor(n_estimators=17, max_features=m, random_state=41)\n",
        "        currrf.fit(train_data, train_sln)\n",
        "        curr_predictions = currrf.predict(test_data)\n",
        "        curr_MSE = metrics.mean_squared_error(test_sln, curr_predictions)\n",
        "        randomforest_MSEs.append(curr_MSE)\n",
        "\n",
        "    plt.suptitle('Random Forest MSE vs. Max Features',fontsize=20)\n",
        "    plt.xlabel('Max Features')\n",
        "    plt.ylabel('Value')\n",
        "    plt.plot(max_features_values,randomforest_MSEs,'ro-', label = \"MSE\")\n",
        "    plt.axis([0,max_features_values[-1]+1,.2,.3])\n",
        "    plt.legend()\n",
        "    plt.show()\n",
        "\n",
        "    print(min(randomforest_MSEs))\n",
        "tune_max_features(range(1,7))"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.24113518626266991\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZEUV9VdxrFb8"
      },
      "source": [
        "The lowest MSE is at 3 features so I will use that earlier in the project. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wAOCA1Q0OWS4"
      },
      "source": [
        "# Bumps in the Road:\n",
        "What challenges did you encounter? How did you overcome these challenges? Did you have to adapt your strategy to account for these challenges? Why or why not?\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T1fpdGUrrk-E"
      },
      "source": [
        "The first challenge I encountered was figuring out which graphs to use. I figured I should use all of them and so I made a code block for each so I wouldnt forget any. I also had a few experiemental problems, where I decided to try using the dummy regressor, which did not work. I then forgot I had used it and none of my datasets worked so I had to reload the Happiness dataset (this took a decent amount of time before I realized). I also sturggled to get the graphs working right for figuring out the right k, and number of trees to use. For K, on the homework the smaller values seemed to work better so I did trial and error for the best one. For trees, I was able to make a graph to see where the best number of trees to use was. \n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qBSirz71PG62"
      },
      "source": [
        "# Conclusions\n",
        "What insights/recommendations do you have? What did you find that was interesting? Which model was your best model, which models didn't work well? Why do you think this is? In general, I want a discussion of your experiment, the results, and what they mean."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GN7vn0D2xjer"
      },
      "source": [
        "Overall, the non-normalized decision forest was the best predictor, it had the lowest MSE (.22) and the highest R^2 (.82). I found it very interesting that the normalized model did not perform better. I do think that because this dataset is always on a 10.0 scale is probably wouldnt do much to change the bias. But I still would have expected otherwise. The model that did the worst for non-normalized was manually calculated knn algorithm. It had an MSE of .3 which is very high compared to the other ones. Between the normalized and the nonnormalized algorithms, the decision tree performed the worst. It had .39 MSE and .68 R^2 for both. This probably means that there arent any predictors that are especially good at predicting the target- Score. So all of the predictors do a good job contriubting to prediciting the happiness score. This would also explain why the knns performed well compared to the decision tree. "
      ]
    }
  ]
}