linear_regression_exercise.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/xiptos/is_notes/blob/main/linear_regression_exercise.ipynb)"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "9c6fc278dd689010"
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Multiple variable linear regression\n",
    "\n",
    "This exercise is an introduction to multiple linear regression, using the wine quality dataset."
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "1e2d22d0-33dd-4ea1-933c-07e24a94dac0"
  },
  {
   "cell_type": "markdown",
   "id": "4b66ef95-99fa-4ab3-9d9e-b8f13a3d4d12",
   "metadata": {},
   "source": [
    "## Dataset\n",
    "\n",
    "Loading the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "!pip install wget\n",
    "!python -m wget -o winequality-red.csv \"https://archive.ics.uci.edu/ml/machine-learning-databases/wine-quality/winequality-red.csv\""
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "bf60f07dfa73b820"
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib as plt\n",
    "import numpy as np\n",
    "wineQ = pd.read_csv(\"winequality-red.csv\")"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "68f89b71-948a-4b56-8d5f-42b4fc1bf84b"
  },
  {
   "cell_type": "markdown",
   "id": "6ee82de3-52bc-44b5-ac6b-3bae646b3d4f",
   "metadata": {},
   "source": [
    "### Understanding of the dataset\n",
    "\n",
    "After loading the dataset, it is necessary to understand what it means and what kind of data is available.\n",
    "\n",
    "Explore the dataset using pandas [`DataFrame`](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.html) [`head()`](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.head.html) function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1e7dd9c2-a616-474b-a4f4-cec02c158556",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Print a sample of the dataset\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83f42563-059b-4f29-b0f0-50f9d679a63a",
   "metadata": {},
   "source": [
    "Also check the [`shape`](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.shape.html) and get some [descriptive statistics](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.describe.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c13fdb00-1637-4955-a0c9-6f97c4a44a40",
   "metadata": {},
   "outputs": [],
   "source": [
    "# check original shape\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "851ef591-c6f9-48f6-a47b-ddac761dfa94",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Descriptive statistics\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8aeac44-50a9-43da-93c2-af849ba334a6",
   "metadata": {},
   "source": [
    "Finally, get information about the dataset using [`info`](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.info.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f1211476-85b5-4b01-bbd6-879c038aba61",
   "metadata": {},
   "outputs": [],
   "source": [
    "# information about the dataset\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e5928e0-b8b9-4cc0-abb0-61274ce1e70d",
   "metadata": {},
   "source": [
    "Let's check for the existance of null values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e4e8dd43-3cb7-432a-b841-3b5ae044e766",
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.heatmap(wineQ.isnull(),yticklabels=False,cbar=False,cmap='viridis')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a565bc2-930c-4cd2-a0d9-21e11eff41ce",
   "metadata": {},
   "source": [
    "Make a representation of the all the relations between all the variables in the dataset.\n",
    "\n",
    "> Hint: [`pairplot`](https://seaborn.pydata.org/generated/seaborn.pairplot.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d5f98f73-910c-4db9-be4a-acd0e0d6ce03",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "50335bb1-3e09-410e-8693-56bee5fc8e54",
   "metadata": {},
   "source": [
    "Also check the distribution of the quality variable using [`displot`](https://seaborn.pydata.org/generated/seaborn.displot.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f0684c61-7286-4692-b9af-c39abe540118",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "ab4d1b2f-924c-4f18-8960-26a85aed7ac2",
   "metadata": {},
   "source": [
    "For better understanding, you can check the relation between two variables here.\n",
    "Try to build a [`scatterplot`](https://seaborn.pydata.org/generated/seaborn.scatterplot.html) chart visualizing the following variables:\n",
    "\n",
    "> alcohol vs. density\n",
    ">\n",
    "> fixed acidity vs. density"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4b7505f1-abd1-4b6a-b3cf-b5c1bcf4b6d9",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "3b851ccd-3fba-4cb5-8569-a20d3dae9f06",
   "metadata": {},
   "source": [
    "Check the correlation between variables using [`corr`](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.corr.html) and ploting them in a [`heatmap`](https://seaborn.pydata.org/generated/seaborn.heatmap.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5aedf7cd-fae8-4a8d-975d-96220da2efa0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "bd87183c-8c71-453a-860b-71452feb1004",
   "metadata": {},
   "source": [
    "## Training the model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cee1360e-2420-4102-ae57-1d72807b7d2d",
   "metadata": {},
   "source": [
    "First, separate the independent variables from the dependent variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eb7daf24-7184-4ce1-9d72-3a7754f80583",
   "metadata": {},
   "outputs": [],
   "source": [
    "x = \n",
    "y = "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6b5753d3-0695-447f-9c00-70a8e4e47d16",
   "metadata": {},
   "source": [
    "Further separate the dataset in two: one for training and one for testing, with the later corresponding to 30% of the total size."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "241ad25a-bfc6-415f-a484-cede1380a65e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "x_train, x_test, y_train, y_test = "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b68472ec-b94c-4a3e-b983-87ade56bb014",
   "metadata": {},
   "source": [
    "Now, perform the training:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1c442a31-0004-47cb-b431-3e1f2d70ef09",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "lm = LinearRegression()\n",
    "# ..."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aea8163b-fe63-4159-a87b-e5a5eb052ee9",
   "metadata": {},
   "source": [
    "Print the intercept and coefficients. What do you think is the shape of the `coef_` property? and the `intercept_`?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5d5510f9-dccb-4e57-abb3-7674d099c0a4",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "a71f9aec-e272-4205-824a-897ac2f11444",
   "metadata": {},
   "source": [
    "Calculate the predictions based on the `x_test` set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9b9eb857-0743-4c8d-a313-f7ea34575ca2",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c9e85c2e-3fd0-479f-982e-b0818f587034",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "plt.scatter(y_test,predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "69ec4e8a-f5e9-402d-bb7a-5817a3ff56b1",
   "metadata": {},
   "source": [
    "## Some quality metrics can also be calculated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a873c441-24fc-4cdf-8486-7c79cdd616b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import metrics\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "602bf1e4-ed6f-4d5d-9470-ef9e647e5095",
   "metadata": {},
   "outputs": [],
   "source": [
    "print('MAE:', metrics.mean_absolute_error(y_test, predictions))\n",
    "print('MSE:', metrics.mean_squared_error(y_test, predictions))\n",
    "print('RMSE:', np.sqrt(metrics.mean_squared_error(y_test, predictions)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "482b8a64-aeeb-4be2-be08-8dca31f67584",
   "metadata": {},
   "outputs": [],
   "source": [
    "print('R sq: ',lm.score(x_train,y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9fa89e73-95c0-492c-998f-1599e424fcbf",
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "print('Correlation: ', math.sqrt(lm.score(x_train,y_train)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "de68d7bb-9277-42f1-b053-3f6f627fba56",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_predicted = lm.predict(x_test)\n",
    "\n",
    "plt.title('Comparison of Y values in test and the Predicted values')\n",
    "plt.ylabel('Test Set')\n",
    "plt.xlabel('Predicted values')\n",
    "plt.scatter(y_predicted, y_test,  color='black')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}