{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import quilt\n",
    "from quilt.data.uciml import iris"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "iris"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = iris.tables.bezdek_iris()\n",
    "trainvecs = train.values[:,:4]\n",
    "labels = train['class'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import svm\n",
    "C = 1.0 # magic regularization parameter\n",
    "models = (svm.SVC(kernel='linear', C=C),\n",
    "          svm.LinearSVC(C=C),\n",
    "          svm.SVC(kernel='rbf', gamma=0.7, C=C),\n",
    "          svm.SVC(kernel='poly', degree=3, C=C))\n",
    "models = (clf.fit(trainvecs, labels) for clf in models)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_score as cvs\n",
    "scores = [cvs(m, trainvecs, labels, cv=5) for m in models]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "[a.mean() for a in scores]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from quilt.data.akarve import pydata_book as pb\n",
    "pb.titanic"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pb.titanic.train()\n",
    "test = pb.titanic.test()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# check for nulls\n",
    "train.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test['IsMale'] = (test['Sex'] == 'male').astype(int)\n",
    "train['IsMale'] = (train['Sex'] == 'male').astype(int)\n",
    "test['NumRelatives'] = test['SibSp'] + test['Parch']\n",
    "train['NumRelatives'] = train['SibSp'] + train['Parch']\n",
    "features = ['Pclass', 'IsMale', 'Age', 'NumRelatives']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# imputation\n",
    "age_median = train['Age'].median()\n",
    "age_mean = train['Age'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "age_mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train['AgeImputeMean'] = train['Age'].fillna(age_mean)\n",
    "test['AgeImputeMean'] = test['Age'].fillna(age_mean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "# select four features we care about\n",
    "features = ['Pclass', 'IsMale', 'NumRelatives', 'AgeImputeMean']\n",
    "# store updated data\n",
    "pb._set(['titanic', 'features'], pd.DataFrame([features]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "features = pb.titanic.features()\n",
    "train = pb.titanic.train()\n",
    "trainsub = train[features.values[0]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainvecs = trainsub.values\n",
    "trainlabels = train['Survived'].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_score as cvs\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "rfc = RandomForestClassifier(max_depth=3, random_state=0)\n",
    "scores = cvs(rfc, trainvecs, trainlabels, cv=5)\n",
    "scores.mean()"
   ]
  }
 ],
 "metadata": {
  "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}