{
 "metadata": {
  "name": "",
  "signature": "sha256:8f2ab6fc66b30e50ac6d8c111699dab1e3580fdf87071476699891b0f2545d30"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Linear Regression using Pandas"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Import what you need\n",
      "import numpy as np\n",
      "import pandas as pd\n",
      "import matplotlib.pyplot as plt"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 84
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Load the data"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Set the data\n",
      "data = pd.DataFrame ({\n",
      "'length' : [94,74,147,58,86,94,63,86,69,72,128,85,82,86,88,72,74,61,90,89,68,76,114,90,78],\n",
      "'weight' : [130,51,640,28,80,110,33,90,36,38,366,84,80,83,70,61,54,44,106,84,39,42,197,102,57]\n",
      "})\n",
      "\n",
      "# create another data frame of log values\n",
      "data_log = np.log(data)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 85
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Build the models"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# ========================\n",
      "# Model for Original Data\n",
      "# ========================\n",
      "\n",
      "# Get the linear models\n",
      "lm_original = np.polyfit(data.length, data.weight, 1)\n",
      "\n",
      "# calculate the y values based on the co-efficients from the model\n",
      "r_x, r_y = zip(*((i, i*lm_original[0] + lm_original[1]) for i in data.length))\n",
      "\n",
      "# Put in to a data frame, to keep is all nice\n",
      "lm_original_plot = pd.DataFrame({\n",
      "'length' : r_x,\n",
      "'weight' : r_y\n",
      "})\n",
      "\n",
      "# ========================\n",
      "# Model for Log Data\n",
      "# ========================\n",
      "\n",
      "# Get the linear models\n",
      "lm_log = np.polyfit(data_log.length, data_log.weight, 1)\n",
      "\n",
      "# calculate the y values based on the co-efficients from the model\n",
      "r_x, r_y = zip(*((i, i*lm_log[0] + lm_log[1]) for i in data_log.length))\n",
      "\n",
      "# Put in to a data frame, to keep is all nice\n",
      "lm_log_plot = pd.DataFrame({\n",
      "'length' : r_x,\n",
      "'weight' : r_y\n",
      "})\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 86
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Plot the models"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# ========================\n",
      "# Plot the data\n",
      "# ========================\n",
      "fig, axes = plt.subplots(nrows=1, ncols=2)\n",
      "\n",
      "# Plot the original data and model\n",
      "data.plot(kind='scatter', color='Blue', x='length', y='weight', ax=axes[0], title='Original Values')\n",
      "lm_original_plot.plot(kind='line', color='Red', x='length', y='weight', ax=axes[0])\n",
      "\n",
      "# Plot the log transformed data and model\n",
      "data_log.plot(kind='scatter', color='Blue', x='length', y='weight', ax=axes[1], title='Log Values')\n",
      "lm_log_plot.plot(kind='line', color='Red', x='length', y='weight', ax=axes[1])\n",
      "\n",
      "plt.show()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 87
    }
   ],
   "metadata": {}
  }
 ]
}