0031 Decision Tree Classifier on Guitar Models.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:bdb34f3de8b7ae52f03159ebba95621b2c59cf5cb898265abd1ff854f6b4fb63"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%matplotlib inline\n",
      "import sys\n",
      "import string\n",
      "import numpy as np\n",
      "import pandas as pd\n",
      "import matplotlib.pyplot as plt\n",
      "import seaborn as sns\n",
      "\n",
      "from sklearn import feature_extraction\n",
      "from sklearn import tree\n",
      "from sklearn.externals.six import StringIO  \n",
      "from IPython.core.display import Image\n",
      "import pydot\n",
      "\n",
      "from __future__ import print_function\n",
      "\n",
      "# turn of data table rendering\n",
      "pd.set_option('display.notebook_repr_html', False)\n",
      "sns.set_palette(['#00A99D', '#F5CA0C', '#B6129F', '#76620C', '#095C57'])\n",
      "sys.version"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 1,
       "text": [
        "'2.7.8 |Anaconda 2.1.0 (64-bit)| (default, Jul  2 2014, 15:12:11) [MSC v.1500 64 bit (AMD64)]'"
       ]
      }
     ],
     "prompt_number": 1
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## What is a Decision Tree?\n",
      "A decision tree is a structure of questions and answers used to separate data points into classes. We can use supervised machine learning to build such a structure from existing data. Decision trees can be used for classification and regression. In this example we focus on classification.\n",
      "\n",
      "### Classification of Guitar Models\n",
      "In the example below we try to predict the class of a guitar, based on its features, using a decision tree. There are only two classes of guitar models in this case, 10 Stratocasters (st) and 16 Les Paul (lp) models. Our feature set contains body material, fretboard, number of frets and kind of pickup elements. Note: this is probably *highly inaccurate* toy-data only created to illustrate a point. You can [download the data set](https://raw.githubusercontent.com/remondo/NoteBooks-Unsupervised-Learning/master/data/guitar-model.csv) from my GitHub repo."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Load the guitar model data set\n",
      "df = pd.read_csv('data\\guitar-model.csv')\n",
      "df"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 2,
       "text": [
        "   model  material fretboard  frets     elements\n",
        "0     st     alder     maple     21   humbuckers\n",
        "1     st     alder     maple     21   humbuckers\n",
        "2     st      lime     maple     22  single coil\n",
        "3     st      lime     maple     22  single coil\n",
        "4     st     alder     maple     24  single coil\n",
        "5     st     alder     maple     24  single coil\n",
        "6     st     alder  rosewood     24  single coil\n",
        "7     st     alder  rosewood     24  single coil\n",
        "8     st     maple  rosewood     24  single coil\n",
        "9     st     maple  rosewood     24  single coil\n",
        "10    lp      lime  rosewood     24   humbuckers\n",
        "11    lp      lime  rosewood     24   humbuckers\n",
        "12    lp     alder  rosewood     24   humbuckers\n",
        "13    lp     alder  rosewood     24   humbuckers\n",
        "14    lp  mahogany  rosewood     24   humbuckers\n",
        "15    lp  mahogany  rosewood     24   humbuckers\n",
        "16    lp  mahogany  rosewood     24   humbuckers\n",
        "17    lp  mahogany  rosewood     22   humbuckers\n",
        "18    lp  mahogany  rosewood     22   humbuckers\n",
        "19    lp     alder     ebony     24   humbuckers\n",
        "20    lp      lime     ebony     22   humbuckers\n",
        "21    lp  mahogany     ebony     22   humbuckers\n",
        "22    lp  mahogany  rosewood     22   humbuckers\n",
        "23    lp  mahogany  rosewood     22   humbuckers\n",
        "24    lp      lime  rosewood     22  single coil\n",
        "25    lp      lime  rosewood     23  single coil"
       ]
      }
     ],
     "prompt_number": 2
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "##Feature Extraction\n",
      "We are confronted with a lot of categorical data, so we need to do some feature extraction first. We use [binary one-hot encoding](http://unsupervised-learning.com/binary-one-hot-encoding-for-machine-learning-in-python/) for this."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Do some feature extracting for\n",
      "cat_columns = ['material', 'fretboard', 'frets', 'elements']\n",
      "cat_dict = df[cat_columns].to_dict(orient='records')\n",
      "\n",
      "vec = feature_extraction.DictVectorizer()\n",
      "cat_vector = vec.fit_transform(cat_dict).toarray()\n",
      "\n",
      "df_vector = pd.DataFrame(cat_vector)\n",
      "vector_columns = vec.get_feature_names()\n",
      "df_vector.columns = vector_columns\n",
      "df_vector.index = df.index\n",
      "\n",
      "df = df.drop(cat_columns, axis=1)\n",
      "df = df.join(df_vector)\n",
      "df.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 3,
       "text": [
        "  model  elements=humbuckers  elements=single coil  fretboard=ebony  \\\n",
        "0    st                    1                     0                0   \n",
        "1    st                    1                     0                0   \n",
        "2    st                    0                     1                0   \n",
        "3    st                    0                     1                0   \n",
        "4    st                    0                     1                0   \n",
        "\n",
        "   fretboard=maple  fretboard=rosewood  frets  material=alder  material=lime  \\\n",
        "0                1                   0     21               1              0   \n",
        "1                1                   0     21               1              0   \n",
        "2                1                   0     22               0              1   \n",
        "3                1                   0     22               0              1   \n",
        "4                1                   0     24               1              0   \n",
        "\n",
        "   material=mahogany  material=maple  \n",
        "0                  0               0  \n",
        "1                  0               0  \n",
        "2                  0               0  \n",
        "3                  0               0  \n",
        "4                  0               0  "
       ]
      }
     ],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Assign an ID to the models\n",
      "df.loc[df.model == 'st','model'] = 0\n",
      "df.loc[df.model == 'lp','model'] = 1\n",
      "df.model.value_counts()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 4,
       "text": [
        "1    16\n",
        "0    10\n",
        "dtype: int64"
       ]
      }
     ],
     "prompt_number": 4
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## Building the Decision Tree Classifier\n",
      "We use Scikit Learn's [DecisionTreeClassifier](http://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeClassifier.html) to construct a decision tree. To choose which feature gives the largest information gain at any given point in the tree, we use the entropy criterion. Entropy is a proportional measure of how pure a set of labels is, where 0.0 is perfectly pure and 1.0 is the largets possible mix of labels."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Split the data set in features and labels\n",
      "features = df.drop(['model'], axis=1)\n",
      "labels = df.model\n",
      "\n",
      "test_features = features[-1:]\n",
      "test_label = labels[-1:]\n",
      "\n",
      "# Train the decision tree based on the entropy criterion\n",
      "clf = tree.DecisionTreeClassifier(criterion='entropy')\n",
      "clf = clf.fit(features[:-1], labels[:-1])\n",
      "clf"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "DecisionTreeClassifier(compute_importances=None, criterion='entropy',\n",
        "            max_depth=None, max_features=None, max_leaf_nodes=None,\n",
        "            min_density=None, min_samples_leaf=1, min_samples_split=2,\n",
        "            random_state=None, splitter='best')"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Visualize the decision tree\n",
      "dot_data = StringIO() \n",
      "tree.export_graphviz(clf, out_file=dot_data, feature_names=features.columns) \n",
      "graph = pydot.graph_from_dot_data(dot_data.getvalue()) \n",
      "Image(graph.create_png())"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "png": "iVBORw0KGgoAAAANSUhEUgAAA1AAAAG1CAYAAAD3DRUpAAAABmJLR0QA/wD/AP+gvaeTAAAgAElE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\nERERmfT/H1AZnpppYAYAAAAASUVORK5CYII=\n",
       "prompt_number": 6,
       "text": [
        "<IPython.core.display.Image at 0x171f2a20>"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Make a prediction with test data\n",
      "pred = clf.predict(test_features)\n",
      "print(features[-1:].T)\n",
      "print('Predicted class:', pred)\n",
      "print('Accurate prediction?', pred[0] == test_label.values[0])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "                      25\n",
        "elements=humbuckers    0\n",
        "elements=single coil   1\n",
        "fretboard=ebony        0\n",
        "fretboard=maple        0\n",
        "fretboard=rosewood     1\n",
        "frets                 23\n",
        "material=alder         0\n",
        "material=lime          1\n",
        "material=mahogany      0\n",
        "material=maple         0\n",
        "Predicted class: [1]\n",
        "Accurate prediction? True\n"
       ]
      }
     ],
     "prompt_number": 7
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "It seems that the 'material' of the guitar body does not play any roll in deciding which label belongs to a given feature set."
     ]
    }
   ],
   "metadata": {}
  }
 ]
}