wine/wine-selection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Wine Selection\n",
    "\n",
    "## Framing\n",
    "I want to buy a fine wine but I have no idea about wine selection.I'm not good at wine tasting.\n",
    "\n",
    "I will use the data and understand what goes into making fine wine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "plt.style.use('ggplot')\n",
    "plt.rcParams['figure.figsize'] = (13,8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"./winequality-red.csv\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Wine Category\n",
    "Let's create a new column 'category' which signifies the category of wine - High (1) or Low (0)\n",
    "\n",
    "Wine with quality > 6 is considered to be High quality, rest are Low quality"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#df.loc[df.b > 0, 'd'] = 1\n",
    "df.loc[df.quality > 5, 'category'] = 1\n",
    "df.loc[df.quality <= 5, 'category'] = 0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the frequency count for each category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.category.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visual Exploration\n",
    "\n",
    "Let's see how the columns are related\n",
    "\n",
    "To start, lets take 2 variables at a time to explore data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Correlation\n",
    "![](../img/corr.svg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "from pandas.tools.plotting import scatter_matrix\n",
    "scatter_matrix(df, figsize=(15,15), diagonal='kde')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Alcohol vs Category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df.plot(x=\"alcohol\", y=\"category\", kind=\"scatter\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Exercise: Volatile Acidity vs Category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3 variable visualization\n",
    "Let's add one more dimension to get more sense what is correlated"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Alcohol vs Volatile Acidity vs Category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "#df.plot(x=\"alcohol\", y=\"volatile acidity\", kind=\"scatter\", c=\"category\")\n",
    "ax = df[df.category == 1].plot(x=\"alcohol\", y=\"volatile acidity\", kind=\"scatter\", color=\"red\", label=\"HIGH\", s=100, alpha=0.5)\n",
    "df[df.category == 0].plot(x=\"alcohol\", y=\"volatile acidity\", kind=\"scatter\", color=\"green\", label=\"LOW\", s=100, alpha=0.5, ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "pd.set_option(\"precision\",3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Time to build a predictive model\n",
    "\n",
    "Let's build a model that can predict the category of wine, given information about alcohol content and volatile acidity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Building a predictive model involves training the model with historical data known as **training data**. Once we have the model trained, the model can predict labels (in this case, the category of wine) for the given features (test data)\n",
    "We have 1600 rows of the wine data, lets split this data into 80:20 ratio as training:testingg data\n",
    "\n",
    "**Why do we need to do this? **\n",
    "\n",
    "We can compare the predicted label with the actual label.\n",
    "By doing this, we can measure how accurate our model is."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_train = df.iloc[:1280,]\n",
    "df_test = df.iloc[1280:,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_train = df_train[\"volatile acidity\"]\n",
    "y_train = df_train[\"category\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_test = df_test[\"volatile acidity\"]\n",
    "y_test = df_test[\"category\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_train = X_train.reshape(X_train.shape[0],1)\n",
    "\n",
    "X_test = X_test.reshape(X_test.shape[0],1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "logistic_model = LogisticRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "logistic_model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "sns.lmplot(data=df, x=\"alcohol\", y=\"category\", logistic=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> It’s a bird… it’s a plane… it… depends on your classifier’s threshold\n",
    "-- [Sancho McCann](https://sanchom.wordpress.com/2011/09/01/precision-recall/)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "predicted = logistic_model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_compare = pd.DataFrame()\n",
    "df_compare[\"actual\"] = y_test\n",
    "df_compare[\"predicted\"] = predicted\n",
    "df_compare[\"volatile acidity\"] = df_test[\"volatile acidity\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "ax=df_compare.plot(x=\"volatile acidity\", y=\"actual\", kind=\"scatter\", color=\"blue\", label=\"actual\")\n",
    "df_compare.plot(x=\"volatile acidity\", y=\"predicted\", kind=\"scatter\", color=\"red\", label=\"predicted\", ax=ax)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's add more features - volatile acidity, sulphates, alcohol to predict the category"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2 variable model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_train = df.iloc[:1280,]\n",
    "df_test = df.iloc[1280:,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_train = df_train[[\"sulphates\", \"alcohol\"]]\n",
    "y_train = df_train[\"category\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_test = df_test[[\"sulphates\", \"alcohol\"]]\n",
    "y_test = df_test[\"category\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "logistic_model = LogisticRegression()\n",
    "\n",
    "logistic_model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "predicted = logistic_model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_compare = pd.DataFrame()\n",
    "df_compare[\"actual\"] = y_test\n",
    "df_compare[\"predicted\"] = predicted\n",
    "df_compare[\"sulphates\"] = df_test[\"sulphates\"]\n",
    "df_compare[\"alcohol\"] = df_test[\"alcohol\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_compare.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "ax = df_compare[df_compare.actual == 1].plot(x=\"alcohol\", y=\"sulphates\", kind=\"scatter\", color=\"red\", label=\"HIGH\", s=100, alpha=0.5)\n",
    "df_compare[df_compare.actual == 0].plot(x=\"alcohol\", y=\"sulphates\", kind=\"scatter\", color=\"green\", label=\"LOW\", s=100, alpha=0.5, ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "ax = df_compare[df_compare.predicted == 1].plot(x=\"alcohol\", y=\"sulphates\", kind=\"scatter\", color=\"red\", label=\"HIGH\", s=100, alpha=0.5)\n",
    "df_compare[df_compare.predicted == 0].plot(x=\"alcohol\", y=\"sulphates\", kind=\"scatter\", color=\"green\", label=\"LOW\", s=100, alpha=0.5, ax=ax)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Accuracy Metrics\n",
    "* AUC\n",
    "* ROC\n",
    "* Misclassification Rate\n",
    "* Confusion Matrix\n",
    "* Precision & Recall"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix\n",
    "\n",
    "![](../img/confusion-matrix.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculate True Positive Rate\n",
    "    TPR = TP / (TP+FN)\n",
    "\n",
    "#### Calculate False Positive Rate\n",
    "    FPR = FP / (FP+TN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Precise & Recall"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### AUC-ROC for the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#ols_auc = metrics.roc_auc_score(df_compare.actual, df_compare.predicted)\n",
    "fpr, tpr, thresholds = metrics.roc_curve(df_compare.actual, df_compare.predicted)\n",
    "plt.plot(fpr, tpr)\n",
    "plt.plot([0,1],[0,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}