ClearML Data and Model Versioning ML project.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Iljf7sDXgGXt"
   },
   "source": [
    "# MLOps\n",
    "\n",
    "**Assignment 1: Data Versioning using clear.ml**\n",
    "\n",
    "+ **Carlson, Nic**\n",
    "+ **Guardi, Tom**\n",
    "+ **Leonard, Andrew**\n",
    "+ **Tyomkin, Lev**\n",
    "\n",
    "**Due date: Thursday, Apr 22, 2021: 6pm**\n",
    "\n",
    "**Objective:** Leverage clear.ml versioning platform using popular ML dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BIcWBCP3gGXx"
   },
   "source": [
    "**Refer to:**\n",
    " - https://allegro.ai/clearml/docs/docs/getting_started/getting_started_clearml_hosted_service.html\n",
    " - https://www.youtube.com/watch?v=lD84X0_0TBE\n",
    " \n",
    "**Steps:**\n",
    " - Step 1: Select popular ML dataset, import and make dataset dirty\n",
    " - Step 2: Initialize clear.ml project, version data, and then clean data and re-version\n",
    " - Step 3: Clean, model, and version\n",
    " - Step 4: Transform, model, and version\n",
    " - Step 5: Compare model and dataset versions in UI\n",
    "\n",
    "**Additional resources on clear.ml:**\n",
    " - Clearml GitHub:\n",
    "    - https://github.com/allegroai/clearml\n",
    "    - https://clearml.ml\n",
    " \n",
    " - Getting started:\n",
    "    - https://allegro.ai/clearml/docs/rst/getting_started/index.html\n",
    "    - Simple sklearn training example: [here](https://github.com/allegroai/clearml/blob/master/examples/frameworks/scikit-learn/sklearn_joblib_example.py)\n",
    "\n",
    " - [Ui walk through](https://allegro.ai/clearml/docs/rst/webapp/index.html)\n",
    "\n",
    " - [Simple code example (uploading data, and using it)](https://github.com/allegroai/events/tree/master/odsc20-east/generic)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zjpPJgutgGX1"
   },
   "source": [
    "#  Step 1: Import and make dataset dirty"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mLtKiIsegGX3"
   },
   "source": [
    "Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "id": "jF53YTVTgGX4"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import sklearn\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BcqQqdXYgGX5"
   },
   "source": [
    "b) Read in the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "hnzUbnVohW_Z",
    "outputId": "d47aa949-af5f-4b8b-b013-5a8070ffd503"
   },
   "outputs": [],
   "source": [
    "# from google.colab import drive\n",
    "# drive.mount('/content/drive')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "id": "pQEwZkhBSzVo"
   },
   "outputs": [],
   "source": [
    "# enter your drive here\n",
    "path = \"2019.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "id": "FXaZRHJzgGX7"
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(path, index_col = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "BDZesgL-PRHL",
    "outputId": "02603dc2-f13d-4a30-f22e-596b4995e934"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Country or region               0\n",
       "Score                           0\n",
       "GDP per capita                  0\n",
       "Social support                  0\n",
       "Healthy life expectancy         0\n",
       "Freedom to make life choices    0\n",
       "Generosity                      0\n",
       "Perceptions of corruption       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 142
    },
    "id": "XpNstmTmgGX8",
    "outputId": "e2f434dc-eb86-4db1-b77b-00c553b90774"
   },
   "outputs": [
    {
     "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>Country or region</th>\n",
       "      <th>Score</th>\n",
       "      <th>GDP per capita</th>\n",
       "      <th>Social support</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",
       "    <tr>\n",
       "      <th>Overall rank</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Finland</td>\n",
       "      <td>7.769</td>\n",
       "      <td>1.340</td>\n",
       "      <td>1.587</td>\n",
       "      <td>0.986</td>\n",
       "      <td>0.596</td>\n",
       "      <td>0.153</td>\n",
       "      <td>0.393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Denmark</td>\n",
       "      <td>7.600</td>\n",
       "      <td>1.383</td>\n",
       "      <td>1.573</td>\n",
       "      <td>0.996</td>\n",
       "      <td>0.592</td>\n",
       "      <td>0.252</td>\n",
       "      <td>0.410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Norway</td>\n",
       "      <td>7.554</td>\n",
       "      <td>1.488</td>\n",
       "      <td>1.582</td>\n",
       "      <td>1.028</td>\n",
       "      <td>0.603</td>\n",
       "      <td>0.271</td>\n",
       "      <td>0.341</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Country or region  Score  GDP per capita  Social support  \\\n",
       "Overall rank                                                            \n",
       "1                      Finland  7.769           1.340           1.587   \n",
       "2                      Denmark  7.600           1.383           1.573   \n",
       "3                       Norway  7.554           1.488           1.582   \n",
       "\n",
       "              Healthy life expectancy  Freedom to make life choices  \\\n",
       "Overall rank                                                          \n",
       "1                               0.986                         0.596   \n",
       "2                               0.996                         0.592   \n",
       "3                               1.028                         0.603   \n",
       "\n",
       "              Generosity  Perceptions of corruption  \n",
       "Overall rank                                         \n",
       "1                  0.153                      0.393  \n",
       "2                  0.252                      0.410  \n",
       "3                  0.271                      0.341  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "s5MdYZbZiP7G"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 156 entries, 1 to 156\n",
      "Data columns (total 8 columns):\n",
      " #   Column                        Non-Null Count  Dtype  \n",
      "---  ------                        --------------  -----  \n",
      " 0   Country or region             156 non-null    object \n",
      " 1   Score                         156 non-null    float64\n",
      " 2   GDP per capita                156 non-null    float64\n",
      " 3   Social support                156 non-null    float64\n",
      " 4   Healthy life expectancy       156 non-null    float64\n",
      " 5   Freedom to make life choices  156 non-null    float64\n",
      " 6   Generosity                    156 non-null    float64\n",
      " 7   Perceptions of corruption     156 non-null    float64\n",
      "dtypes: float64(7), object(1)\n",
      "memory usage: 11.0+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0rQBr7NZgGX9"
   },
   "source": [
    "Make data dirty"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 142
    },
    "id": "mcCEfsjckG31",
    "outputId": "51e4e495-f0f5-4418-e89c-9cb652bbeba7"
   },
   "outputs": [
    {
     "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>Country or region</th>\n",
       "      <th>Score</th>\n",
       "      <th>GDP per capita</th>\n",
       "      <th>Social support</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",
       "    <tr>\n",
       "      <th>Overall rank</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Finland</td>\n",
       "      <td>7.769</td>\n",
       "      <td>-1.340</td>\n",
       "      <td>-1.587</td>\n",
       "      <td>-0.986</td>\n",
       "      <td>-0.596</td>\n",
       "      <td>-0.153</td>\n",
       "      <td>-0.393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Denmark</td>\n",
       "      <td>7.600</td>\n",
       "      <td>1.383</td>\n",
       "      <td>1.573</td>\n",
       "      <td>0.996</td>\n",
       "      <td>0.592</td>\n",
       "      <td>0.252</td>\n",
       "      <td>0.410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Norway</td>\n",
       "      <td>7.554</td>\n",
       "      <td>1.488</td>\n",
       "      <td>1.582</td>\n",
       "      <td>1.028</td>\n",
       "      <td>0.603</td>\n",
       "      <td>0.271</td>\n",
       "      <td>0.341</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Country or region  Score  GDP per capita  Social support  \\\n",
       "Overall rank                                                            \n",
       "1                      Finland  7.769          -1.340          -1.587   \n",
       "2                      Denmark  7.600           1.383           1.573   \n",
       "3                       Norway  7.554           1.488           1.582   \n",
       "\n",
       "              Healthy life expectancy  Freedom to make life choices  \\\n",
       "Overall rank                                                          \n",
       "1                              -0.986                        -0.596   \n",
       "2                               0.996                         0.592   \n",
       "3                               1.028                         0.603   \n",
       "\n",
       "              Generosity  Perceptions of corruption  \n",
       "Overall rank                                         \n",
       "1                 -0.153                     -0.393  \n",
       "2                  0.252                      0.410  \n",
       "3                  0.271                      0.341  "
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfDirty = df.copy()\n",
    "dfDirty.iloc[0,2:] = dfDirty.iloc[0,2:]*-1\n",
    "dfDirty.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "47xrcCvVMlDN"
   },
   "source": [
    "Save dirty dataset to local data folder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_folder = \"data/\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "id": "uSrETLubMj7U"
   },
   "outputs": [],
   "source": [
    "dfDirty.to_csv(dataset_folder + \"happiness_data.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cquN64r2gGYC"
   },
   "source": [
    "# Step 2: Initialize clear.ml project with dirty data, model, and version"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VkwyyXpcgGYE"
   },
   "source": [
    "Start using clear.ml with the dirty dataset and version the dataset.\n",
    "\n",
    "If notebook is idle too long you might need to uncomment and install"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "rzO-Y2JnlIX1",
    "outputId": "efab66b2-4ad9-44ff-8b54-ac0a9d6a9945"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting clearml\n",
      "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/38/01/37831856a7cf71b35dab8c915e2ff0d61776f851673f17021e74470c9fd1/clearml-0.17.5-py2.py3-none-any.whl (888kB)\n",
      "\u001b[K     |████████████████████████████████| 890kB 5.8MB/s \n",
      "\u001b[?25hRequirement already satisfied: numpy>=1.10 in /usr/local/lib/python3.7/dist-packages (from clearml) (1.19.5)\n",
      "Requirement already satisfied: requests>=2.20.0 in /usr/local/lib/python3.7/dist-packages (from clearml) (2.23.0)\n",
      "Requirement already satisfied: future>=0.16.0 in /usr/local/lib/python3.7/dist-packages (from clearml) (0.16.0)\n",
      "Requirement already satisfied: jsonschema>=2.6.0 in /usr/local/lib/python3.7/dist-packages (from clearml) (2.6.0)\n",
      "Collecting pyjwt<3.0.0,>=1.6.4\n",
      "  Downloading https://files.pythonhosted.org/packages/b4/9b/8850f99027ed029af6828199cc87179eaccbbf1f9e6e373e7f0177d32dad/PyJWT-2.0.1-py3-none-any.whl\n",
      "Collecting humanfriendly>=2.1\n",
      "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/93/66/363d01a81da2108a5cf446daf619779f06d49a0c4426dd02b40734f10e2f/humanfriendly-9.1-py2.py3-none-any.whl (86kB)\n",
      "\u001b[K     |████████████████████████████████| 92kB 3.9MB/s \n",
      "\u001b[?25hCollecting pathlib2>=2.3.0\n",
      "  Downloading https://files.pythonhosted.org/packages/e9/45/9c82d3666af4ef9f221cbb954e1d77ddbb513faf552aea6df5f37f1a4859/pathlib2-2.3.5-py2.py3-none-any.whl\n",
      "Requirement already satisfied: Pillow>=4.1.1 in /usr/local/lib/python3.7/dist-packages (from clearml) (7.1.2)\n",
      "Requirement already satisfied: psutil>=3.4.2 in /usr/local/lib/python3.7/dist-packages (from clearml) (5.4.8)\n",
      "Requirement already satisfied: pyparsing>=2.0.3 in /usr/local/lib/python3.7/dist-packages (from clearml) (2.4.7)\n",
      "Requirement already satisfied: python-dateutil>=2.6.1 in /usr/local/lib/python3.7/dist-packages (from clearml) (2.8.1)\n",
      "Requirement already satisfied: PyYAML>=3.12 in /usr/local/lib/python3.7/dist-packages (from clearml) (3.13)\n",
      "Requirement already satisfied: six>=1.11.0 in /usr/local/lib/python3.7/dist-packages (from clearml) (1.15.0)\n",
      "Requirement already satisfied: urllib3>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from clearml) (1.24.3)\n",
      "Requirement already satisfied: attrs>=18.0 in /usr/local/lib/python3.7/dist-packages (from clearml) (20.3.0)\n",
      "Collecting furl>=2.0.0\n",
      "  Downloading https://files.pythonhosted.org/packages/12/18/b29367947b32b510cbbbfa86164929ceed069ff020f84a6dc780df5d6ba1/furl-2.1.2-py2.py3-none-any.whl\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests>=2.20.0->clearml) (2020.12.5)\n",
      "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests>=2.20.0->clearml) (2.10)\n",
      "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests>=2.20.0->clearml) (3.0.4)\n",
      "Collecting orderedmultidict>=1.0.1\n",
      "  Downloading https://files.pythonhosted.org/packages/04/16/5e95c70bda8fe6ea715005c0db8e602400bdba50ae3c72cb380eba551289/orderedmultidict-1.0.1-py2.py3-none-any.whl\n",
      "Installing collected packages: pyjwt, humanfriendly, pathlib2, orderedmultidict, furl, clearml\n",
      "Successfully installed clearml-0.17.5 furl-2.1.2 humanfriendly-9.1 orderedmultidict-1.0.1 pathlib2-2.3.5 pyjwt-2.0.1\n"
     ]
    }
   ],
   "source": [
    "# !pip install clearml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "noAd8Auz-7as"
   },
   "source": [
    "# Input credentials from Nic's workspace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "phhgLJJPNWmp"
   },
   "outputs": [],
   "source": [
    "# # Configuration\n",
    "# api {\n",
    "#     # Nic Carlson's workspace\n",
    "#     web_server: https://app.community.clear.ml\n",
    "#     api_server: https://api.community.clear.ml\n",
    "#     files_server: https://files.community.clear.ml\n",
    "#     credentials {\n",
    "#         \"access_key\" = \"########\"\n",
    "#         \"secret_key\" = \"########\"\n",
    "#     }\n",
    "# }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "T-LLRFxileL-",
    "outputId": "91578435-ba25-432d-b14d-dfe114930ad6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ClearML SDK setup process\n",
      "Configuration file already exists: C:\\Users\\carlson nic\\clearml.conf\n",
      "Leaving setup, feel free to edit the configuration file.\n"
     ]
    }
   ],
   "source": [
    "!clearml-init"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "syzLYewTgGYL",
    "outputId": "6e771297-8984-444b-a448-3bd420d2ee33"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ClearML Task: created new task id=321177277e4f46ab91b5d60de7540ad9\n",
      "ClearML results page: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/321177277e4f46ab91b5d60de7540ad9/output/log\n"
     ]
    }
   ],
   "source": [
    "# Establish first task\n",
    "from clearml import Task\n",
    "task = Task.init(project_name=\"assignment1\", task_name=\"dirty_data_preProcessEDA\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FyCL7lmjMTBf"
   },
   "source": [
    "Upload dirty dataset to ClearML (i.e. create version 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "hlyqOKBtgGYO",
    "outputId": "bafffe11-8b8e-4be5-a0b7-f624e7e20c07"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:09:10,668 - clearml.Task - INFO - No repository found, storing script code instead\n",
      "ClearML Monitor: GPU monitoring failed getting GPU reading, switching off GPU monitoring\n",
      "clearml-data - Dataset Management & Versioning CLI\n",
      "Creating a new dataset:\n",
      "New dataset created id=cf15b353fd5140bc9f83e7b70d5332cb\n",
      "clearml-data - Dataset Management & Versioning CLI\n",
      "Syncing dataset id cf15b353fd5140bc9f83e7b70d5332cb to local folder data/happiness_data.csv\n",
      "Sync completed: 0 files removed, 1 added / modified\n",
      "Finalizing dataset\n",
      "Pending uploads, starting dataset upload to https://files.community.clear.ml\n",
      "Uploading compressed dataset changes (1 files, total 4.63 KB) to https://files.community.clear.ml\n",
      "Upload completed (4.63 KB)\n",
      "2021-04-17 18:09:43,887 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:09:44,107 - clearml.Task - INFO - Finished uploading\n",
      "Dataset closed and finalized\n"
     ]
    }
   ],
   "source": [
    "# Upload data to ClearML\n",
    "!clearml-data create --project assignment1 --name 'dirty_dataset'\n",
    "\n",
    "# Sync local data folder with ClearML\n",
    "!clearml-data sync --folder data/happiness_data.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zD-YuzG4gGYQ"
   },
   "source": [
    "Run EDA (exploratory data analysis) of the dirty dataset - submit the plots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 371
    },
    "id": "HXTe-hJ9gGYS",
    "outputId": "013d2391-ed2c-4e71-9eda-355b79bdb257"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# GDP is a strong predictor\n",
    "plt.hist(dfDirty['GDP per capita'], bins=8)\n",
    "plt.title(\"Hist Dirty\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 371
    },
    "id": "51AIhN6GOFpq",
    "outputId": "1d54dd36-e73d-4052-e8bf-e3af9537af9c"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(dfDirty['GDP per capita'],dfDirty['Healthy life expectancy'])\n",
    "plt.title(\"Scatter Dirty\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 340
    },
    "id": "Gku6Z3Z_OTmA",
    "outputId": "3386bf6a-baee-49b2-f5bd-67d0c0a9052f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:09:45,602 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:09:59,088 - clearml.Task - INFO - Finished uploading\n"
     ]
    }
   ],
   "source": [
    "task.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "d7SfBHlUgGYT"
   },
   "source": [
    "### Dirty Model\n",
    "Use the dirty dataset to build a baseline machine learning model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "E_1pnmgurxQX"
   },
   "source": [
    "Train Test Split\n",
    "\n",
    "__Target Variable: Healthy Life Expectancy__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "id": "5HR5EeScS45m"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ClearML Task: created new task id=703e16249122496a84ddfa5c42d31d00\n",
      "ClearML results page: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/703e16249122496a84ddfa5c42d31d00/output/log\n"
     ]
    }
   ],
   "source": [
    "# Establish task\n",
    "task = Task.init(project_name=\"assignment1\", task_name=\"training dirty\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "id": "KOWihREEgGYV"
   },
   "outputs": [],
   "source": [
    "# predictors\n",
    "X = dfDirty[['GDP per capita',\n",
    "             'Social support',\n",
    "             'Freedom to make life choices', \n",
    "             'Generosity', \n",
    "             'Perceptions of corruption']]\n",
    "# target\n",
    "y = dfDirty['Healthy life expectancy']\n",
    "\n",
    "# train test split\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X, y, test_size=0.2, random_state=426)\n",
    "\n",
    "# upload processed data\n",
    "# print('Uploading process dataset')\n",
    "# task.upload_artifact('X_train', X_train)\n",
    "# task.upload_artifact('X_test', X_test)\n",
    "# task.upload_artifact('y_train', y_train)\n",
    "# task.upload_artifact('y_test', y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "fBzzD5asrqr7"
   },
   "source": [
    "Random Forest Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "id": "Gf2gpGeDgGYW"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(max_depth=2, random_state=0)"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "rf_dirty = RandomForestRegressor(max_depth=2, random_state=0)\n",
    "rf_dirty.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "id": "uj_k1ww7A-po"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:10:04,014 - clearml.Task - INFO - No repository found, storing script code instead\n",
      "ClearML Monitor: GPU monitoring failed getting GPU reading, switching off GPU monitoring\n",
      "model trained & stored\n"
     ]
    }
   ],
   "source": [
    "import joblib\n",
    "joblib.dump(rf_dirty, 'rf_dirty.pkl', compress=True)\n",
    "\n",
    "loaded_model = joblib.load('rf_dirty.pkl')\n",
    "result = loaded_model.score(X_test, y_test)\n",
    "\n",
    "print('model trained & stored')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "id": "EwiT-zFYxfzK"
   },
   "outputs": [],
   "source": [
    "y_pred = rf_dirty.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YhvZ9NnrzyxW"
   },
   "source": [
    "Results\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "id": "LHEqdQtkxeT3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 0.07702010004921575\n",
      "Mean Squared Error: 0.011539677070834004\n",
      "Root Mean Squared Error: 0.10742288895218748\n"
     ]
    }
   ],
   "source": [
    "# Evaluating the Algorithm\n",
    "from sklearn import metrics\n",
    "print('Mean Absolute Error:', metrics.mean_absolute_error(y_test, y_pred))  \n",
    "print('Mean Squared Error:', metrics.mean_squared_error(y_test, y_pred))  \n",
    "print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "id": "_KqYnaWSx93v"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Results on Test\n",
    "plt.scatter(y_pred, y_test)\n",
    "plt.title(\"RF Dirty Test Scatter\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "X75rfV2Syiff"
   },
   "source": [
    "Training results look overfit and you can see our Finland outlier clearly"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "id": "r-4yTOvQyQoI"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Results on Train\n",
    "plt.scatter(rf_dirty.predict(X_train), y_train)\n",
    "plt.title(\"RF Dirty Predictions\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.barh(width = rf_dirty.feature_importances_, y=X.columns)\n",
    "plt.title(\"Feature Importance Dirty\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "id": "duYres2dTjuj"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:10:11,331 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:10:22,367 - clearml.Task - INFO - Finished uploading\n"
     ]
    }
   ],
   "source": [
    "task.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8ChoAXScgGYW"
   },
   "source": [
    "# Step 3: Clean, model, and version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 340
    },
    "id": "eukzTKPYudc4",
    "outputId": "db248cd2-cf40-4ee9-8517-8a12ddcb2c23"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ClearML Task: created new task id=c96383281b9040bf9b002f374da6c51e\n",
      "ClearML results page: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/c96383281b9040bf9b002f374da6c51e/output/log\n"
     ]
    }
   ],
   "source": [
    "task = Task.init(project_name=\"assignment1\", task_name=\"clean_data_preProcessEDA\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "id": "WDQVpp50gGYY"
   },
   "outputs": [],
   "source": [
    "# Reverse dirtying steps above\n",
    "dfClean = dfDirty.copy()\n",
    "dfClean.iloc[0,2:] = dfClean.iloc[0,2:]*-1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "id": "AZ_rR2ZPSJep"
   },
   "outputs": [],
   "source": [
    "dfClean.to_csv(dataset_folder+\"/happiness_data.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "b43gFhWvgGYZ"
   },
   "source": [
    "Update the dataset version to go from dirty to clean dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "id": "jdxITS-ygGYa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:10:26,289 - clearml.Task - INFO - No repository found, storing script code instead\n",
      "ClearML Monitor: GPU monitoring failed getting GPU reading, switching off GPU monitoring\n",
      "Generating SHA2 hash for 2 files\n",
      "Hash generation completed\n",
      "Uploading compressed dataset changes (1 files, total 358 bytes) to https://files.community.clear.ml\n",
      "Upload completed (358 bytes)\n",
      "2021-04-17 18:10:37,704 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:10:37,898 - clearml.Task - INFO - Finished uploading\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Do clear.ml magic e.g. version\n",
    "from clearml import Dataset\n",
    "dataset = Dataset.get(dataset_id = 'a7dbe610552441c1af83a3716994216e')\n",
    "\n",
    "new_dataset = Dataset.create(dataset_project = 'assignment1', dataset_name = 'clean_dataset', parent_datasets=[dataset])\n",
    "new_dataset.sync_folder(local_path=dataset_folder)\n",
    "new_dataset.upload()\n",
    "new_dataset.finalize()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1Y-9herMgGYf"
   },
   "source": [
    "Run EDA (exploratory data analysis) of the clean dataset - submit the plots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 371
    },
    "id": "ftupMddV2YEk",
    "outputId": "6b2ec0bc-1b85-4066-9f79-16e3e561a1ac"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# GDP is a strong predictor\n",
    "plt.hist(df['GDP per capita'], bins=8)\n",
    "plt.title(\"Hist Clean\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 371
    },
    "id": "pLvqRiL7GZfo",
    "outputId": "dbe28786-3947-43cd-a605-1eba64cc6fbf"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(df['GDP per capita'],df['Healthy life expectancy'])\n",
    "plt.title(\"Scatter Clean\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "id": "raL8RwNVPE9D"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:10:38,752 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:10:40,977 - clearml.Task - INFO - Finished uploading\n"
     ]
    }
   ],
   "source": [
    "task.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dNhFAjXZgGYk"
   },
   "source": [
    "Build a machine learning model with this \"new\" clean dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "id": "mXjuWhWTHQl4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ClearML Task: created new task id=afbf86e003904852a9a09c78d87e2cc4\n",
      "ClearML results page: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/afbf86e003904852a9a09c78d87e2cc4/output/log\n"
     ]
    }
   ],
   "source": [
    "# Establish task\n",
    "task = Task.init(project_name=\"assignment1\", task_name=\"training clean\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "id": "4cJi6muv12ww"
   },
   "outputs": [],
   "source": [
    "# predictors\n",
    "X = dfClean[['GDP per capita',\n",
    "             'Social support',\n",
    "             'Freedom to make life choices', \n",
    "             'Generosity', \n",
    "             'Perceptions of corruption']]\n",
    "# target\n",
    "y = dfClean['Healthy life expectancy']\n",
    "\n",
    "# train test split\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X, y, test_size=0.2, random_state=426)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YfKdLeM212xA"
   },
   "source": [
    "Random Forest Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "id": "iSwXo-Z_12xA"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(max_depth=2, random_state=0)"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "rf_clean = RandomForestRegressor(max_depth=2, random_state=0)\n",
    "rf_clean.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "id": "XntOTsZfHHZc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:10:46,563 - clearml.Task - INFO - No repository found, storing script code instead\n",
      "ClearML Monitor: GPU monitoring failed getting GPU reading, switching off GPU monitoring\n",
      "model trained & stored\n"
     ]
    }
   ],
   "source": [
    "import joblib\n",
    "joblib.dump(rf_clean, 'rf_clean.pkl', compress=True)\n",
    "\n",
    "loaded_model = joblib.load('rf_clean.pkl')\n",
    "result = loaded_model.score(X_test, y_test)\n",
    "\n",
    "print('model trained & stored')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "id": "6aUzD6x212xB"
   },
   "outputs": [],
   "source": [
    "y_pred = rf_clean.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9IuFsR5f12xB"
   },
   "source": [
    "Results\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "id": "UTQ2UPs_12xB"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 0.07253981653567454\n",
      "Mean Squared Error: 0.010195385417798429\n",
      "Root Mean Squared Error: 0.10097220121299935\n"
     ]
    }
   ],
   "source": [
    "# Evaluating the Algorithm\n",
    "from sklearn import metrics\n",
    "print('Mean Absolute Error:', metrics.mean_absolute_error(y_test, y_pred))  \n",
    "print('Mean Squared Error:', metrics.mean_squared_error(y_test, y_pred))  \n",
    "print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "id": "euZzmk-K12xC"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Results on Test\n",
    "plt.scatter(y_pred, y_test)\n",
    "plt.title(\"Scatter Cleaned TEST\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "S6dH9JKU12xC"
   },
   "source": [
    "Training results look overfit and clustered"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "id": "eJ1xrH4-12xC"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Results on Train\n",
    "plt.scatter(rf_clean.predict(X_train), y_train)\n",
    "plt.title(\"Scatter Cleaned TRAIN\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "id": "O9P4j9Q112xC"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.barh(width = rf_clean.feature_importances_, y=X.columns)\n",
    "plt.title(\"Feature Importance Cleaned\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = rf_clean.predict(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "id": "qt371qHKBrsp"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 0.08626760794632085\n",
      "Mean Squared Error: 0.013644154248769632\n",
      "Root Mean Squared Error: 0.1168081942706488\n"
     ]
    }
   ],
   "source": [
    "# Evaluating the Algorithm - on Train\n",
    "from sklearn import metrics\n",
    "print('Mean Absolute Error:', metrics.mean_absolute_error(y_train, y_pred))  \n",
    "print('Mean Squared Error:', metrics.mean_squared_error(y_train, y_pred))  \n",
    "print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_train, y_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "id": "ACBgSc1YY81I"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:10:51,806 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:11:04,982 - clearml.Task - INFO - Finished uploading\n"
     ]
    }
   ],
   "source": [
    "task.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HyB8iU5NgGYm"
   },
   "source": [
    "# Step 4: Transform, model, and version"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xApSbuXKgGYo"
   },
   "source": [
    "Transform the clean dataset using aggregation - call this the transformed dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "id": "AxVpB9hcumVf"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ClearML Task: created new task id=c5d2e0926aeb40408bf3b6891e73abd5\n",
      "ClearML results page: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/c5d2e0926aeb40408bf3b6891e73abd5/output/log\n"
     ]
    }
   ],
   "source": [
    "task = Task.init(project_name=\"assignment1\", task_name=\"transformed_data_preProcessEDA\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "id": "CG-GR4jOgGYB"
   },
   "outputs": [],
   "source": [
    "df['Social Generosity'] = df['Social support'] * df['Generosity']\n",
    "df['Score'] = np.log(df['Score'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "id": "WIjG-droVMAn"
   },
   "outputs": [],
   "source": [
    "df.to_csv(dataset_folder+'/happiness_data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dHQoFq_AgGYr"
   },
   "source": [
    "Update the dataset version with the transformed dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "id": "3dVXFRXRgGYp"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:11:08,037 - clearml.Task - INFO - No repository found, storing script code instead\n",
      "ClearML Monitor: GPU monitoring failed getting GPU reading, switching off GPU monitoring\n",
      "Generating SHA2 hash for 2 files\n",
      "Hash generation completed\n",
      "Uploading compressed dataset changes (2 files, total 6.87 KB) to https://files.community.clear.ml\n",
      "Upload completed (6.87 KB)\n",
      "2021-04-17 18:11:21,386 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:11:21,622 - clearml.Task - INFO - Finished uploading\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Do clear.ml magic e.g. version\n",
    "new_dataset = Dataset.create(dataset_project = 'assignment1', dataset_name = 'transformed_dataset', parent_datasets=[dataset])\n",
    "new_dataset.sync_folder(local_path=dataset_folder)\n",
    "new_dataset.upload()\n",
    "new_dataset.finalize()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "leo5xdZpgGYy"
   },
   "source": [
    "Run EDA (exploratory data analysis) of the transformed dataset - submit the plots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 371
    },
    "id": "ifY7ep-XaRsC",
    "outputId": "6b2ec0bc-1b85-4066-9f79-16e3e561a1ac"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# GDP is a strong predictor\n",
    "plt.hist(df['GDP per capita'], bins=8)\n",
    "plt.title(\"Hist TRANSFORMED\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 371
    },
    "id": "lv4sWr5daRsE",
    "outputId": "dbe28786-3947-43cd-a605-1eba64cc6fbf"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(df['GDP per capita'],df['Healthy life expectancy'])\n",
    "plt.title(\"Scatter TRANSFORMED\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "id": "jlWvKIJgaRsE"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:11:22,517 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:11:28,090 - clearml.Task - INFO - Finished uploading\n"
     ]
    }
   ],
   "source": [
    "task.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "85iY6ZVFgGY0"
   },
   "source": [
    "Build a machine learning model with the transformed dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "id": "DSkrvP5iapc5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ClearML Task: created new task id=118ef3701f25408dae0dd12d7f0e04a0\n",
      "ClearML results page: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/118ef3701f25408dae0dd12d7f0e04a0/output/log\n"
     ]
    }
   ],
   "source": [
    "# Establish task\n",
    "task = Task.init(project_name=\"assignment1\", task_name=\"training transformed\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "id": "_Q9FY3o855vo"
   },
   "outputs": [],
   "source": [
    "# predictors\n",
    "X = df[['GDP per capita',\n",
    "             'Social support',\n",
    "             'Freedom to make life choices', \n",
    "             'Generosity', \n",
    "             'Perceptions of corruption',\n",
    "             'Social Generosity']]\n",
    "# target\n",
    "y = df['Healthy life expectancy']\n",
    "\n",
    "# train test split\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X, y, test_size=0.2, random_state=426)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "o7AbbMz155vv"
   },
   "source": [
    "Random Forest Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "id": "qgYwAwhP55vw"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(max_depth=2, random_state=0)"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "rf_trans = RandomForestRegressor(max_depth=2, random_state=0)\n",
    "rf_trans.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:11:31,504 - clearml.Task - INFO - No repository found, storing script code instead\n",
      "ClearML Monitor: GPU monitoring failed getting GPU reading, switching off GPU monitoring\n",
      "model trained & stored\n"
     ]
    }
   ],
   "source": [
    "joblib.dump(rf_trans, 'rf_trans.pkl', compress=True)\n",
    "\n",
    "loaded_model = joblib.load('rf_trans.pkl')\n",
    "result = loaded_model.score(X_test, y_test)\n",
    "\n",
    "print('model trained & stored')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "id": "MQC2H94F55vx"
   },
   "outputs": [],
   "source": [
    "y_pred = rf_trans.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "I15s5G-655vy"
   },
   "source": [
    "Results\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "id": "jWmS2gRP55vy"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 0.07328917782726302\n",
      "Mean Squared Error: 0.010390462790936347\n",
      "Root Mean Squared Error: 0.10193361953220512\n"
     ]
    }
   ],
   "source": [
    "# Evaluating the Algorithm\n",
    "from sklearn import metrics\n",
    "print('Mean Absolute Error:', metrics.mean_absolute_error(y_test, y_pred))  \n",
    "print('Mean Squared Error:', metrics.mean_squared_error(y_test, y_pred))  \n",
    "print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "id": "KfBVkG6d55vy"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Results on Test\n",
    "plt.scatter(y_pred, y_test)\n",
    "plt.title(\"Scatter Transformed TEST\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jL5qFsKu55vz"
   },
   "source": [
    "Training results look overfit and clustered"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "id": "KmphpM9a55vz"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/d3fzzAAAACXBIWXMAAAsTAAALEwEAmpwYAAAjmElEQVR4nO3df5xddX3n8dc7k4lO1DIgqSvDj0SLQZECMgJdui2gNkCrROoq/qiVR1uWXem2rpuSthahtUu6PKw/VluWuq7LagUqdhYUG/sQrS01wqQJPwLEpqAkE61BCSqJMEk++8c5E27unHPvuXfuj3PPfT8fj3kk95xzz/2eO3c+93s+31+KCMzMbPAt6ncBzMysMxzQzcwqwgHdzKwiHNDNzCrCAd3MrCIc0M3MKsIB3ayOEv9b0uOS7up3eepJ+qakV/e7HFY+DugVJelnJf2jpCckfV/SnZJeucBzvkPSP9Rt+4Sk9y2stIec7wuSfpT+zEp6uubxdZ16nSZ+FngNcHREnN6j11ywZu+dpLMlHUgf/1DSVkmXZJznHZJC0hvrtp8taUfN469I+rGkY2q2vVrSN7t6oZbLAb2CJP0E8DngfwBHABPA1cBT/SxXFkmLax9HxPkR8dyIeC7wKeC/zz2OiMvyntdhxwHfjIgnW31il8vVUMH3bme6/yeAdwF/IWll3al+Ffh++m8zTwJ/0KFLsAVyQK+mlwBExKcjYn9E7I2IL0bEvXMHSPoNSQ+mNbUHJL0i3b5W0r/UbH99uv2lwHXAz6Q1vN2SLgXeCvxOuu229NijJN0iaZekRyT955rXvUrSZyR9UtIPgHcUvai01vhOSf8M/HO67UOStkv6gaSNkv5d3WvdLOmG9Hq2SJqs2X+FpJma2uqrJP0a8LGa67y65v3alt7t3CrpqLxyzdVkJf2OpO9K+rak1ZIukPSN9By/V/P8RTXv+/fSMh9Rs/9XJH0r3ff7Rd+vRiJxO0ng/uma1zoO+HngUmCVpBc0OdWHgTdL+qlOlMsWKCL8U7EfktrX94D/A5wPHF63/98DM8ArAQE/BRxXs+8oki/7N5HUwF6Y7nsH8A915/oE8L6ax4uAjcCVwBLgRcDDwKp0/1XALLA6PXaswXXUnzuAvyW56xhLt70NeD6wGHg38B3g2TWv9WPgAmAEuAbYkO5bCWwHjkofLwdenHWdwLnAY8ArgGeR3Pl8Na9cwNnAvvQ9GAV+A9gF/CXwPODEtFwvSp//28AG4Oj0/P8T+HS672XAj4CfS/f9aXruVzf5DBzy3qXbzgZ21PyeXgccAE6tOeYPgLvS/98H/Jes56ePvwL8elqmT6bbXk1yd9P3v4Nh/HENvYIi4gckeeAA/gLYldYq52pbv05yO353JLZFxLfS5/5VROyMiAMRcRNJTbiVPPIrgWUR8YcR8XREPJyW4eKaY74WEVPpa+xt8fKuiYjvzz0vIj4ZEd+LiH0R8X6SoFebQviHiLg9IvYD/xc4Od2+Pz32ZZJGI+KbEfEvOa/5VuDjEfFPEfEU8LskNfjleeUi+dL644iYBW4EjgQ+FBE/jIgtwBaeqRn/B+D3I2JHev6rgDek6Zs3AJ+LiK+m+/6AJAi36yhJu4G9wF+TBOxNNfvfTvLFQ/pvkbTLNcBrJZ24gHJZBzigV1REPBgR74iIo4GXk9S6P5juPgbIDF6S3i5pc5pS2Z0+98gWXvo40qBRc47fA2pv3be3dDGHOuS5kt6dpo6eSF/rsLryfqfm/3uAZ0taHBHbSGrGVwHflXRjbRqlzlHAt+YeRMSPSO6AJvLKBXwv/RKBJHgC/GvN/r3Ac9P/Hwf8dc379SDJF84L0tc+eO5I8vrfyylnETsjYpzkLu7DJHcfAEg6C1hB8gUESUA/SdIpjU4YEbuAjwB/uIByWQc4oA+BiHiI5Bb85emm7cCL649L86d/AVwOPD/9w7+fJC0DSY1/3unrHm8HHomI8Zqf50XEBQ2e04qDz03z5VcAbyRJK40DT9SUt/GJIv4yIn6WJKAG8Cc5h+5Mj5l73eeQpHlmssrVhu3A+XXv2bMjYgb4NskX8NxrL01fe0HS2v4VJAF7dbr5V0neu82SvgN8Pd3+9gKnvBY4BzhtoWWz9jmgV5CkE9Ka69Hp42OAN5PkaSFp9Puvkk5T4qfSYP4cksC0K33eJTzzJQBJDfNoSUvqtr2o5vFdwA/SBscxSSOSXq4FdpnM8TySfPIuYLGkK0lqnk1JWinpXEnPIsln7yWpFWf5S+ASSaekx/834OsR8c2FXkDqOuCP098BkpZJujDd9xngl5R0Q11CUgvuyN9tRDwNvB+4UtKzSb4YLwVOqfn5TeCtatJ7JyJ2p+f6nU6UzdrjgF5NPwTOAL4u6UmSQH4/SaMhEfFXwB+TBKofAlPAERHxAMkf5ddIAvVJwJ01572DJPf7HUmPpdv+F0keerekqTTN8FqSYPAISWPix0hSIZ22HvgC8A2SlMiPKZ7OeRawLi3fd4CfJEkNzRMRXyLJXd9CUmN+MYe2CSzUh4BbgS9K+iHJ7+uM9LW3AO8k+V19G3gc2JFznnZ8HDiWpDF8L3BDRHxn7ofk9zsCnFfwOvK+FK0HFOEFLszMqsA1dDOzinBANzOrCAd0M7OKcEA3M6uIvk0kdOSRR8by5cv79fJmZgNp48aNj0XEsqx9fQvoy5cvZ3p6ul8vb2Y2kCR9K2+fUy5mZhXhgG5mVhEO6GZmFeGAbmZWEQ7oZmYV0bdeLmZmw2Zq0wzXrt/Kzt17OWp8jDWrVrL61InmTyzIAd3MrAemNs3wu5+9j72zyYSUM7v38rufvQ+gY0HdKRczsx64dv3Wg8F8zt7Z/Vy7fmvHXsMB3cysB3buzl4+N297OxzQzcx64KjxsZa2t8MB3cysB845IXP6ldzt7XBANzPrgS8/tKul7e1wQDcz6wHn0M3MKsI5dDOzkpnaNMNZ6+5gxdrPc9a6O5jaNFPoeWtWrWRsdOSQbWOjI6xZtbJjZfPAIjOzghYyOGhuv0eKmpmVQKPBQUUC8+pTJzoawOs55WJmVlAvGjYXwgHdzKygXjRsLoQDuplZQWtWrWR0kQ7ZNrpIHW3YXAgHdDOzVqjJ4z5yo6iZWUHXrt/K7P44ZNvs/ijcKOr50M3MSqJoo2hW4Aa6Ph+6A7qZWUFHjY8xkxHUaxtF8/qqP3t00YK6PBbhHLqZWUHLn5/dm6V2e15f9cf3zGY+N+sLol0O6GZmBW14+PGm2/vZJ90B3cysoP0RTbf3s0+6A7qZWQflTcLVCw7oZmYdtPrUCa656CQmxscQMDE+xjUXndST13YvFzOzgiZyerlM1KVZsibh+u2bNnezaECBGrqkj0v6rqT7c/ZL0oclbZN0r6RXdL6YZtYN7czt3e584FWQlU4RnV0XdCGKpFw+AZzXYP/5wPHpz6XAny+8WGbWbXP9pWd27yV4pr90owDdznOqZPWpE/zyaROHjPYP4JaNM6V4D5oG9Ij4KvD9BodcCNwQiQ3AuKQXdqqAZtYdjeb27uRzqubLD+2ivq9LWd6DTjSKTgDbax7vSLfNI+lSSdOSpnft6txK12bWunbm9i77fOC9UOb3oBMBPWuusczOmhFxfURMRsTksmXlyDmZDat25vYu+3zgvVDm96ATAX0HcEzN46OBnR04r5l1UTuLFvdioeOyK/N70Ilui7cCl0u6ETgDeCIivt2B85pZG4pO0drOosW9WOi47Np9D0akzJGmI+rchOqKnKGsBw+QPg2cDRwJ/CvwXmAUICKukyTgIyQ9YfYAl0TEdLMXnpycjOnppoeZWQvqZ/qDpPZ4zUUnDVXQLaP3TN3HJzc8Om/72848lvetLj7wSNLGiJjM2te0hh4Rb26yP4B3Fi6NmXXNQlelt+6ZC9qf/vp29kcwIvHmM45pKZg345GiZhVS5h4YlgT1Tgbwep7LxaxCytwDw7rPAd2sQsrcA8O6zykXswpxL5Th5oBuVjFZM/3ZcHBANxtyRfutW/nfK+fQzYbY1KYZ1nzmnkNmT1zzmXtKMXNg2WTNNPmumzbznqn7+l20gxzQzYbY1bdtYXb/oYMLZ/cHV9+2pU8lKq+sPv4BfGrDo6X5AnRANxtij++ZbWn7MMvryx9QiqlzwTl0s4FU9lxuFR2Vs/wclGfglmvoZgOmk6sGjY+NtrR9mK1ZtTJzrnAoz8AtB3SzAdOJVYPm1gXdvTc7tfJLJy9s0bEqrju6+tQJ3nrmsfOCepkGbjnlYjZgFjpfS9aMjPW+/FD7K4rVn3/uDgIY+LTQ+1afxORxR5Q23eWAbjYg5vLmeRNez932N8uvZ9Xw67WTE5573aw8s2d87A0HdLMBUKRW/eRT+3jP1H3csnGmYe24SLBuNSdcpHxlaThciLLffTiHbjYAitSqd++d5VMbHm2aXx9f2rjBs52ccJHylaXhcCE60X7RTa6hmw2AorXbvHTM3POnNs3wox/vy33+RJs54WblK1PD4UKUfb55B3SzAdCoD3QRc7Xya9dvZfbA/LA/PjbK5vf+QlfK1+6XRBnlXWdZ7j6ccjEbAFnznLeytPDc0sF5NckncrovFpU3D/sH33QKd649txLBHMo/37wDutkAWH3qBNdcdBIT42OIpNb71jOPnRdc8swF7G6taJRVviouTF3263TKxSpnWIbFTx53xCF9ohdJ7I/sLPph6cjPNatWzuuNklXDbOc9HJZ52Mt8nYqcD0C3TU5OxvT0dF9e26orq/vc2OhIqWpR7ShyXe+Zuo9Pbng08/lLRxfxwB+df/BctcH6nBOW8eWHdh3yuLbrY9ZrDbt+VhokbYyIyax9rqFbpTTqVjbIwajIdTUa3bln9sDB/9fWMLP6VX9qw6Pzess0eg+H5Y5oTpn7ojugW6WUvVsZtBYApzbNcNWtW3LnXKm9rlausdGozmZdH+vPU9bg1i1lrjS4UdQqpVuNfp3SykyJU5tmWPNX9+QGczj0uhpd4+E1g4lqy9CKrPOXfaBNN5R5Ct1CAV3SeZK2StomaW3G/sMk3SbpHklbJF3S+aKaNVf2bmWtBMC8PuNz6q/rnBOWZR4n4L2vPbFhGZoRZL6Hg3BH1ElTm2ZKPYVu04AuaQT4KHA+8DLgzZJeVnfYO4EHIuJk4Gzg/ZKWdLisZk2VvVtZKwGwWVCsv668HHr9V0I7wTbITqGU/Y6o0/ImR8v7wuu1Ijn004FtEfEwgKQbgQuBB2qOCeB5kgQ8F/g+kD++2KyL2u1W1ovGvVZGGjYbfVlftkaB+rdv2sy167eyZtXKhucdyen6OJEToIt2g6yKvPct7wuv14qkXCaA7TWPd6Tban0EeCmwE7gP+K2IOFB3DJIulTQtaXrXrvbnWzbrtE6uAtRIKymhNatWMrpo/g3+6IhYs2rlvEUkDmuyytDcNZ1zwrLcUZ3vf+PJLaWsyn5H1Gkjyk645G3vtSI19KyS1n+FrwI2A+cCLwb+VtLfR8QPDnlSxPXA9ZD0Q2+5tGZd0queC3PnKnInMLettpfL4UtH+cWffuG8ni8zu/cyOtI8qOyd3c+XH9rFNRed1HTO9KJ3KmUeaNNpeQO38rb3WpGAvgM4pubx0SQ18VqXAOsiGaW0TdIjwAnAXR0ppVmX9bJxr5UAWH9so3nHZ/cHz1q8iKf2zbs5PsTO3XsblmGYAnSrJnLSVXkpqV4rknK5Gzhe0oq0ofNi4Na6Yx4FXgUg6QXASuDhThbUrJsGpXGvWQ+VZsEcyndNg6TsvaiaBvSI2AdcDqwHHgRujogtki6TdFl62B8B/1bSfcCXgCsi4rFuFdqs08r+hzpnoXcMZbymQVL2NgPP5WKWGoQh7Getu6PtedGrNC/5MPNcLmYFlCl3nPflktVNsBlPrDU8HNDNCupVDb7I/Cjvvvme3J4V42OjPOdZi0t9p2Hd4ZSL9d0gpDrypq/95dMmDpl6thNlz0urTIyPcefacwFYsfbzuZNoffBNp5Tu/bPOccrFSmtQZuvL66deO9Vsp8pepAtl3mjP8bHRUr1v1lsO6NZXZZ6KtFZekG1l3vCiikwPkDfk/qrXnTjvec00W/CijHdMls0BvYsGIZXQb+0O6On1e9to/pN6C+1aWGR+lFZGnDaSdYdUu+pRWe+YLJsDepcMSiqh31qZrGpOP97brCArsheD6MSCy9A8WHeiV06RqXTr7zpcUSkvB/QuGZRUQr+1M1tfP97brCCbt/ZmJwbu9KoLZdG7ibnjXFEpNwf0Lhm2if/b1U7qoJvvbaPaZ1aQnTzuiIGurRZNJc3ddbiiUm4O6F3STiphWLVaG+3We9tO7bNMg5HaUWSgUu1dhysq5eY1RbtkUOYGGUTdem+HcX3MrLlJ3nbmsblzlQzKJGbDyjX0LulULwSbr1vv7bDWPlu5yxi2FYoGjQN6Fw367XiZdeO9dZqsOVdUys0B3UqpH13jXPssxhWV8nJAt9LpV9e4KtQ+3Ud8uDmgW+n0s2vcINc+3UfcHNCtdMrWOFnGWm9WmdxH3BzQrXTK1DhZxlpvXpny+pJXvZeOPcP90K10ytSHvx9906c2zXDWujtYsfbznLXuDqY2zRQq04iUeT730hkerqFb6ZSpcbLX6Z8idwR5r70/grHREffSGWIO6H1UxtxsWZSlcbLX6Z8iefC8Mk3U5NL9mRpODuh9UsbcrM3X677pRe4IGpWpLF+E1h/OoffJMM4bMoiy5jqpnduk04rMldLrMtngcA29T8rWNc/y9bLWW/SOwDVxy+KA3mFF8+Jl6ppn5VGmBmEbPA7oHdRKXtzzhlge176tXYVy6JLOk7RV0jZJa3OOOVvSZklbJP1dZ4s5GFrJizsPamad1rSGLmkE+CjwGmAHcLekWyPigZpjxoE/A86LiEcl/WSXyltqrebFXRMzs04qUkM/HdgWEQ9HxNPAjcCFdce8BfhsRDwKEBHf7WwxB4NXczGzfiqSQ58Attc83gGcUXfMS4BRSV8Bngd8KCJuqD+RpEuBSwGOPfbYdspbOrWNoIeNjTI6Imb3x8H9zoubWa8UCehZE0RE3ePFwGnAq4Ax4GuSNkTENw55UsT1wPUAk5OT9ecYOPWNoLv3zjK6SBy+dJTde2bdQ8E6zqOLrZEiAX0HcEzN46OBnRnHPBYRTwJPSvoqcDLwDfqsm38AWY2gsweCpUsWs+nKX+jIa5jNadSLCtzV0YoF9LuB4yWtAGaAi0ly5rX+H/ARSYuBJSQpmQ90sqDt6Pbweg8Osl7K60V19W1b+PHsAU8jYc0bRSNiH3A5sB54ELg5IrZIukzSZekxDwJ/A9wL3AV8LCLu716xi+n28Ho3glov5VUUHt8z62kkDCg4sCgibgdur9t2Xd3ja4FrO1e09tSmWPKS9J2qQVdtcJDzs+WWN7o4j+8Uh0+lJueaS7HMNAjm0LkadJUGB9W/d3O37fWLK1j/5C38MT42mnm87xSHT6WG/melWOp1ugZdlcFBXo+y/PLmeQEqdado7atUQG90iykolEYY1rSDG3gHw1wFYu5z+q6bNnPU+Bi/fNoEX35o19B9bu1QlQroeTnGw5eOFupGOMyLTnj2x8GR9Tm9ZePMwKb7rHMqlUNfs2oloyPzx0H96Mf7CuWCh3nRiTItzGyNDfPn1BqrVEBffeoEz1ky/6Zj9kAU+rAPc9qhSg28VTfMn1NrrFIpF4An9s5mbi/yYR/2tENVGnirbtg/p5avUjV0yP9Qjy/N7toFSU7yrHV3MLN777yJa5x2sLJxeszyVC6gt5pHr+1/DYfOOua0g5WR02OWRxH9mfRwcnIypqenu3LuU67+IrszUi8T42PcufbcQ7bN1czrCfjAm04BPOmRmZWHpI0RMZm1r3I5dGieRy8yPUCAJz0ys4FSuZQLNJ40q+j0AOBJj8xssFQyoDdqNCoyPUAz7h5m3TTXSL9i7ec5a90dnk/HCqtkQG/UaFQ0GHvSI+sHT5JmC1HJHDrk96luND3A0iWLPemR9dVCJ0kb1rmILFHZgJ5latMMTz61b972sdER3vvaE3M/+P4DsV5ZyCjQYZ6LyBJDE9DrP+xzDl862jCYe/Sk9dJCRoF6CmSrZA49S15j6NIli/1ht9JYyChQz/Fila6h93I5OrNOyFvEokilw3O8WGUDel6Kpd5hOT1ZzPql3TRf1da4tdZVMqBPbZrh3Tffw/4C0xrM7j/QgxKZdd9CavdWDZUL6HM18yLBHODJpxc2yMjdxKxM3Ig/3CrXKNrOSNB2B214EIiZlUnlAno7jZztzs3ipcDMrEwqF9DzWvQXzZ8i/aB2e7q4m5iZlUmhgC7pPElbJW2TtLbBca+UtF/SGzpXxNbk9eP90zeewuE5qxa1262r0ayOZma91rRRVNII8FHgNcAO4G5Jt0bEAxnH/QmwvhsFzdKoQTJveye7dbmbmJmVSZFeLqcD2yLiYQBJNwIXAg/UHfebwC3AKztawhzN5q3IaunvVLeu2i+S8aWjPGvxIp7YO+teLmbWV0UC+gSwvebxDuCM2gMkTQCvB86lRwG93XkrFtqtq/6L5PE9s4yNjvCBN53iQG5mfVUkh57VnFjfyfuDwBUR0bC/oKRLJU1Lmt61a1fBIs43tWkmc4gzJA2S3VwgwD1bzKysitTQdwDH1Dw+GthZd8wkcKMkgCOBCyTti4ip2oMi4nrgekgWiW6nwHM15Dxjo4t4102bD37jdHoKUfdsMbOyKlJDvxs4XtIKSUuAi4Fbaw+IiBURsTwilgOfAf5TfTDvlEYDh0YXiT2zB+bdPnSyBu2eLWZWVk0DekTsAy4n6b3yIHBzRGyRdJmky7pdwHqNasKLGnQ271QNeiHTm5qZdVOhuVwi4nbg9rpt1+Uc+46FFytf3hShAE/ty59oa5HE1KaZBaddPAGSmZXVwE3OldX3u4j9ER3LpXsCJDMro4Eb+r/61Amuueiktp7r3ihmVmUDF9AhCeoTbTZCujeKmVXVQAZ0yG6cLMK9UcysqgY2oM+lXibGxxAwMT7G28489uDj8bFRRkfm93p58ql9nq/czCpp4BpFazVrnJzaNMPVt23h8T2zB7ft3jvb0YFGZmZlMbA19CJWnzrB0iXzv7PcOGpmVVTpgA4eqm9mw2OgUy5Z6udIP2xslN17Z+cd58ZRM6uaSgX0rDnSR0fE6CIxe+CZGV48VN/MqqgSAX2uVp41JcDs/uDwpaMsXbLYQ/XNrNIGPqDX18qz7N4zy6Yrf6GHpTIz672BbxRtNJ3uHOfLzWwYDHxAb9ZbxflyMxsWA59yaTSd7kSP8uX1PWucozezfhj4gJ41ne7Y6AjXXHRST4JqVs8aj0Q1s34Y+JRL1pwuvQrm4EWjzaw8Br6GDv1dcMIjUc2sLAa+ht5vh42NZm53zxoz67VK1NCz9KKhcmrTDE8+vW/e9tFFcs8aM+u5Sgb0XjVUXrt+K7P7Y9725z57sRtEzaznKplyufq2LT1pqMzLk+/eM38yMDOzbqtcQJ/aNHPIgha1Ot1QmZcnd/7czPqhcgG9US2804E2a11Tj0w1s36pXA69US2804F2Lk/uUaJmVgaVC+h5UwGMj412JdD2sw+8mVmtQikXSedJ2ippm6S1GfvfKune9OcfJZ3c+aIWk5cGuep1J/apRGZmvdG0hi5pBPgo8BpgB3C3pFsj4oGawx4Bfj4iHpd0PnA9cEY3CtyM0yBmNqyKpFxOB7ZFxMMAkm4ELgQOBvSI+Mea4zcAR3eykEXVDyb6wJtOcSA3s6FRJOUyAWyvebwj3Zbn14AvZO2QdKmkaUnTu3btKl7KAuYGE83s3kvwzGCiqU0zHX0dM7OyKhLQlbFt/vBIQNI5JAH9iqz9EXF9RExGxOSyZcuKl7IAz3poZsOuSMplB3BMzeOjgZ31B0n6aeBjwPkR8b3OFK84z3poZsOuSEC/Gzhe0gpgBrgYeEvtAZKOBT4L/EpEfKPjpSwgr7viUeNjXlHIzIZC05RLROwDLgfWAw8CN0fEFkmXSbosPexK4PnAn0naLGm6ayXOkddd8ZwTljm3bmZDQRGZ6fCum5ycjOnpzsb9rJr4teu3ZtbcJ8bHuHPtuR19fTOzbpO0MSIms/ZVaqRo1qjNd920OfNY59bNrGoqNzlXPc+IaGbDovIB3TMimtmwqFTKJYunAjCzYVH5gA6eEdHMhkPlUy5mZsPCAd3MrCIc0M3MKmKgcugewm9mlm9gAvrc9LhzMyrODeEHHNTNzBiglIunxzUza2xgArqnxzUza2xgArqH8JuZNTYwAd1D+M3MGhuYRlEP4Tcza2xgAjp4CL+ZWSMDk3IxM7PGHNDNzCrCAd3MrCIc0M3MKsIB3cysIhzQzcwqwgHdzKwiHNDNzCrCAd3MrCIc0M3MKqJQQJd0nqStkrZJWpuxX5I+nO6/V9IrOl/U4TK1aYaz1t3BirWf56x1dzC1aabfRTKzkms6l4ukEeCjwGuAHcDdkm6NiAdqDjsfOD79OQP48/Rfa4NXZzKzdhSpoZ8ObIuIhyPiaeBG4MK6Yy4EbojEBmBc0gs7XNah4dWZzKwdRQL6BLC95vGOdFurxyDpUknTkqZ37drValmHhldnMrN2FAnoytgWbRxDRFwfEZMRMbls2bIi5RtKXp3JzNpRJKDvAI6peXw0sLONY6wgr85kZu0oEtDvBo6XtELSEuBi4Na6Y24F3p72djkTeCIivt3hsg6N1adOcM1FJzExPoaAifExrrnoJDeImllDTXu5RMQ+SZcD64ER4OMRsUXSZen+64DbgQuAbcAe4JLuFXk4eHUmM2tVoSXoIuJ2kqBdu+26mv8H8M7OFs3MzFrhkaJmZhXhgG5mVhEO6GZmFeGAbmZWEUraM/vwwtIu4FtdOPWRwGNdOG8/VOVafB3lU5VrGcbrOC4iMkdm9i2gd4uk6YiY7Hc5OqEq1+LrKJ+qXIuv41BOuZiZVYQDuplZRVQxoF/f7wJ0UFWuxddRPlW5Fl9Hjcrl0M3MhlUVa+hmZkPJAd3MrCIGNqA3W7i65rhXStov6Q29LF9RBRbgPlvSE5I2pz9X9qOcRRT5naTXs1nSFkl/1+syFlHgd7Km5vdxf/r5OqIfZW2kwHUcJuk2Sfekv49SzpJa4DoOl/TX6QL1d0l6eT/K2Yykj0v6rqT7c/ZL0ofT67xX0itafpGIGLgfkml8/wV4EbAEuAd4Wc5xd5DMFPmGfpe7nesAzgY+1++yduhaxoEHgGPTxz/Z73K3+9mqOf61wB39Lnebv4/fA/4k/f8y4PvAkn6XvY3ruBZ4b/r/E4Av9bvcOdfyc8ArgPtz9l8AfIFkBbgzga+3+hqDWkMvsnA1wG8CtwDf7WXhWlD0OgZBkWt5C/DZiHgUICLK+Htp9XfyZuDTPSlZa4pcRwDPkyTguSQBfV9vi9lUket4GfAlgIh4CFgu6QW9LWZzEfFVkvc4z4XADZHYAIxLemErrzGoAb3potSSJoDXA9dRXoUW1wZ+Jr0t/oKkE3tTtJYVuZaXAIdL+oqkjZLe3rPSFVf0d4KkpcB5JJWGsilyHR8BXkqyXOR9wG9FxIHeFK+wItdxD3ARgKTTgeNIlsEcNIU/e3kKLXBRQkUWpf4gcEVE7E8qIKVU5Dr+iWTuhh9JugCYAo7vdsHaUORaFgOnAa8CxoCvSdoQEd/oduFaUGjB89RrgTsjolGtq1+KXMcqYDNwLvBi4G8l/X1E/KDLZWtFketYB3xI0maSL6ZNlO9Oo4hWPnuZBjWgF1mUehK4MQ3mRwIXSNoXEVM9KWExTa+j9o8rIm6X9GeSjoyIsk1IVHQx8cci4kngSUlfBU4GyhTQW1nw/GLKmW6BYtdxCbAukgTuNkmPkOSg7+pNEQsp+jdyCSQNi8Aj6c+gaeWzl63fDQVtNi4sBh4GVvBMQ8mJDY7/BOVsFG16HcC/4ZkBYKcDj849LtNPwWt5KUmuczGwFLgfeHm/y97OZws4jCQf+px+l3kBv48/B65K//8CYAY4st9lb+M6xkkbc4HfIMlD973sOdeznPxG0V/k0EbRu1o9/0DW0KPYwtWlV/A63gD8R0n7gL3AxZH+9sukyLVExIOS/ga4FzgAfCwiMrtw9UsLn63XA1+M5G6jdApexx8Bn5B0H0kQuSJKdudX8DpeCtwgaT9JL6pf61uBG5D0aZJea0dK2gG8FxiFg9dxO0lPl23AHtK7jpZeo4SxwczM2jCovVzMzKyOA7qZWUU4oJuZVYQDuplZRTigm5lVhAO6mVlFOKCbmVXE/wen9KPFLWR9mAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Results on Train\n",
    "plt.scatter(rf_trans.predict(X_train), y_train)\n",
    "plt.title(\"Scatter Transformed TRAIN\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "id": "aQ1dbj-755vz"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.barh(width = rf_trans.feature_importances_, y=X.columns)\n",
    "plt.title(\"Feature Importance Transformed\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "id": "pPvFIHhLay0A"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-17 18:12:50,192 - clearml.Task - INFO - Waiting to finish uploads\n",
      "2021-04-17 18:12:50,591 - clearml.Task - INFO - Finished uploading\n"
     ]
    }
   ],
   "source": [
    "task.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "e_0S77QTgGZF"
   },
   "source": [
    "# Step 5: Compare model and dataset versions in UI"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compare and comment the accuracy of the models using the dirty, clean and transformed datasets."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Transformed:\n",
    "- Mean Absolute Error: 0.07328917782726302\n",
    "- Mean Squared Error: 0.010390462790936347\n",
    "- Root Mean Squared Error: 0.10193361953220512\n",
    "\n",
    "\n",
    "Clean:\n",
    "- Mean Absolute Error: 0.07253981653567454\n",
    "- Mean Squared Error: 0.010195385417798429\n",
    "- Root Mean Squared Error: 0.10097220121299935\n",
    "\n",
    "Dirty:\n",
    "- Mean Absolute Error: 0.07702010004921575\n",
    "- Mean Squared Error: 0.011539677070834004\n",
    "- Root Mean Squared Error: 0.10742288895218748\n",
    "\n",
    "Based on our results, we can observe a clear performance improvement of our Cleaned dataset over our Dirty dataset. This is evident by lower Mean Absolute Error metrics from 0.077 to 0.072, lower Mean Squared Error metrics from 0.012 to 0.01, and consequently lower RMSE from 0.107 to 0.101. Having lower scores of MSE means our Clean Trained predictions are closer to actual values and the same came be inferred for lower MAE as a negatrively-oriented accuracy metric. \n",
    "\n",
    "Our Transformed Dataset predictions actually saw decreased accuracy as MAE, MSE, and RMSE all went slightly higher compared to Clean (while remaining an improvement over our Dirty data). This suggests that feature engineered transformations like 'Score' and 'Social Generosity' did not improve the accuracy of our model predictions. Further testing should be administered to identify more optimal transformations."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Submit the models and the comparison:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![newplot (5).png](data:image/png;base64,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)![newplot (4).png](data:image/png;base64,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)![newplot (3).png](data:image/png;base64,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)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on the EDA of our Dirty dataset, our manually generated outlier set comprised of negative values is evident in the Predictions scatterplot."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![newplot (5).png](data:image/png;base64,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)![newplot (3).png](data:image/png;base64,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)![newplot (4).png](data:image/png;base64,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)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The difference between the cleaned and dirty data is eveident in the scatterplots, removing the outliers previously entered for the dirty set. Visible overfitting of our Training data is illustrated by the clustering effect in the Training scatter. This effect drove our decision to plot  GDP in our EDA - in an effort to see if that clustering was reflected in our feature importance histograms and how it related to the clustering of countries by GDP."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![newplot (5).png](data:image/png;base64,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)![newplot (4).png](data:image/png;base64,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)![newplot (3).png](data:image/png;base64,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)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The transformed version of our Clean data largely resembles the scatterplot of our clean set in trend/spread; however, the noticable difference here is the inclusion of new features evident in our Feature Importance plot. Our engineered feature \"social generosity\" did not change the overall feature importance breakdown and GDP remained dominant."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The full breakdown of our Models/Comparison can be refenced in our ClearML Project Folder comprised of the following Experiment List: \n",
    "\n",
    "training transformed: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/118ef3701f25408dae0dd12d7f0e04a0/output/execution\n",
    "transformed_data_preProcessEDA: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/c5d2e0926aeb40408bf3b6891e73abd5/output/execution\n",
    "transformed_dataset: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/b4d7c1e7d24840708ba1583effff33a8/output/execution\n",
    "training clean: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/afbf86e003904852a9a09c78d87e2cc4/output/execution\n",
    "clean_data_preProcessEDA: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/c96383281b9040bf9b002f374da6c51e/output/execution\n",
    "clean_dataset: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/f346e109ca884ca58a8f4ae0b8c914a1/output/execution\n",
    "training dirty: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/703e16249122496a84ddfa5c42d31d00/output/execution\n",
    "dirty_data_preProcessEDA: https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/321177277e4f46ab91b5d60de7540ad9/output/execution\n",
    "'dirty_dataset': https://app.community.clear.ml/projects/ef0b38af172e49578627be10b55c08bc/experiments/cf15b353fd5140bc9f83e7b70d5332cb/output/execution"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Assignment1.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "384px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "position": {
    "height": "830.969px",
    "left": "983px",
    "right": "20px",
    "top": "73px",
    "width": "594.938px"
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}