03_l1_normalization.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import tensorflow as tf \n",
    "import numpy as np \n",
    "import matplotlib.pyplot as plt\n",
    "from scipy import stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import normalize, minmax_scale"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv('datasets/dataset2.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    157\n",
       "1    262\n",
       "2    272\n",
       "3    223\n",
       "4    159\n",
       "5    153\n",
       "6    247\n",
       "7    259\n",
       "8    224\n",
       "9    142\n",
       "Name: average_montly_hours, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['average_montly_hours'][:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "hours = df['average_montly_hours'].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "## Normalization using scikit L1-norm:\n",
    "It is also known as least absolute deviations (LAD), least absolute errors (LAE). It is basically minimizing the sum of the absolute differences (S) between the target value (Yi) and the estimated values (f(xi))\n",
    "<img src=\"images/l1-norm-formula.png\">\n",
    "\n",
    "To understand easily, its just adding all the values in the array and dividing each of it using the sum"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "result = normalize(df['average_montly_hours'].astype(float).values.reshape(1,-1), norm='l1', axis=1).reshape(-1,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  5.20634019e-05],\n",
       "       [  8.68828746e-05],\n",
       "       [  9.01990148e-05],\n",
       "       ..., \n",
       "       [  4.74208056e-05],\n",
       "       [  9.28519270e-05],\n",
       "       [  5.23950160e-05]])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DescribeResult(nobs=14999, minmax=(array([  3.18349464e-05]), array([ 0.0001028])), mean=array([  6.66711114e-05]), variance=array([  2.74294286e-10]), skewness=array([ 0.0528367]), kurtosis=array([-1.13500325]))"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.describe(result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}