{ "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "path_data = '../../../data/'\n", "\n", "import numpy as np\n", "import pandas as pd\n", "\n", "%matplotlib inline\n", "import matplotlib.pyplot as plt\n", "plt.style.use('fivethirtyeight')\n", "\n", "import warnings\n", "warnings.filterwarnings('ignore')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Prediction ###\n", "\n", "An important aspect of data science is to find out what data can tell us about the future. What do data about climate and pollution say about temperatures a few decades from now? Based on a person's internet profile, which websites are likely to interest them? How can a patient's medical history be used to judge how well he or she will respond to a treatment?\n", "\n", "To answer such questions, data scientists have developed methods for making *predictions*. In this chapter we will study one of the most commonly used ways of predicting the value of one variable based on the value of another." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The foundations of the method were laid by [Sir Francis Galton](https://en.wikipedia.org/wiki/Francis_Galton). As we saw in Section 7.1, Galton studied how physical characteristics are passed down from one generation to the next. Among his best known work is the prediction of the heights of adults based on the heights of their parents. We have studied the dataset that Galton collected for this. The table `heights` contains his data on the midparent height and child's height (all in inches) for a population of 934 adult \"children\"." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "# Galton's data on heights of parents and their adult children\n", "galton = pd.read_csv(path_data + 'galton.csv')\n", "heights = pd.DataFrame(\n", " {'MidParent':galton['midparentHeight'],\n", " 'Child':galton['childHeight']}\n", " )" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
MidParentChild
075.4373.2
175.4369.2
275.4369.0
375.4369.0
473.6673.5
.........
92966.6464.0
93066.6462.0
93166.6461.0
93265.2766.5
93365.2757.0
\n", "

934 rows × 2 columns

\n", "
" ], "text/plain": [ " MidParent Child\n", "0 75.43 73.2\n", "1 75.43 69.2\n", "2 75.43 69.0\n", "3 75.43 69.0\n", "4 73.66 73.5\n", ".. ... ...\n", "929 66.64 64.0\n", "930 66.64 62.0\n", "931 66.64 61.0\n", "932 65.27 66.5\n", "933 65.27 57.0\n", "\n", "[934 rows x 2 columns]" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "heights" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "heights.plot.scatter('MidParent', 'Child')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The primary reason for collecting the data was to be able to predict the adult height of a child born to parents similar to those in the dataset. We made these predictions in Section 7.1, after noticing the positive association between the two variables. \n", "\n", "Our approach was to base the prediction on all the points that correspond to a midparent height of around the midparent height of the new person. To do this, we wrote a function called `predict_child` which takes a midparent height as its argument and returns the average height of all the children who had midparent heights within half an inch of the argument." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "def predict_child(mpht):\n", " \"\"\"Return a prediction of the height of a child \n", " whose parents have a midparent height of mpht.\n", " \n", " The prediction is the average height of the children \n", " whose midparent height is in the range mpht plus or minus 0.5 inches.\n", " \"\"\"\n", " \n", " close_points = heights[(heights['MidParent'] > mpht-0.5) & (heights['MidParent'] < mpht + 0.5)]\n", " \n", " return close_points['Child'].mean() " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We applied the function to the column of `Midparent` heights, visualized our results." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
MidParentChildPrediction
075.4373.270.100000
175.4369.270.100000
275.4369.070.100000
375.4369.070.100000
473.6673.570.415789
............
92966.6464.065.156579
93066.6462.065.156579
93166.6461.065.156579
93265.2766.564.229630
93365.2757.064.229630
\n", "

934 rows × 3 columns

\n", "
" ], "text/plain": [ " MidParent Child Prediction\n", "0 75.43 73.2 70.100000\n", "1 75.43 69.2 70.100000\n", "2 75.43 69.0 70.100000\n", "3 75.43 69.0 70.100000\n", "4 73.66 73.5 70.415789\n", ".. ... ... ...\n", "929 66.64 64.0 65.156579\n", "930 66.64 62.0 65.156579\n", "931 66.64 61.0 65.156579\n", "932 65.27 66.5 64.229630\n", "933 65.27 57.0 64.229630\n", "\n", "[934 rows x 3 columns]" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Apply predict_child to all the midparent heights\n", "\n", "heights_with_predictions = heights.copy()\n", "\n", "heights_with_predictions['Prediction'] = heights['MidParent'].map(predict_child)\n", "\n", "heights_with_predictions" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Draw the original scatter plot along with the predicted values\n", "\n", "fig, ax = plt.subplots(figsize=(8,5))\n", "\n", "ax.scatter(heights_with_predictions['MidParent'], \n", " heights_with_predictions['Child'], \n", " label='Child', \n", " color='darkblue')\n", "\n", "ax.scatter(heights_with_predictions['MidParent'], \n", " heights_with_predictions['Prediction'], \n", " label='Prediction', \n", " color='gold')\n", "\n", "x_label = 'MidParent'\n", "\n", "y_label = ''\n", "\n", "y_vals = ax.get_yticks()\n", "\n", "plt.ylabel(y_label)\n", "\n", "plt.xlabel(x_label)\n", "\n", "ax.legend(bbox_to_anchor=(1.04,1), loc=\"upper left\", frameon=False)\n", "\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The prediction at a given midparent height lies roughly at the center of the vertical strip of points at the given height. This method of prediction is called *regression.* Later in this chapter we will see where this term came from. We will also see whether we can avoid our arbitrary definitions of \"closeness\" being \"within 0.5 inches\". But first we will develop a measure that can be used in many settings to decide how good one variable will be as a predictor of another." ] } ], "metadata": { "anaconda-cloud": {}, "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": "3.6.12" } }, "nbformat": 4, "nbformat_minor": 1 }