model10

model10_GMM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model10: GMM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## A. Functions\n",
    "\n",
    "There have four different functions.\n",
    "\n",
    "* Data reader: Read data from file.\n",
    "* Feature functions(private): Functions which extract features are placed in here. It means that if you make a specific feature function, you can add the one into here.\n",
    "* Feature function(public): We can use only this function for feature extraction.\n",
    "* Utility functions: All the funtions except functions which are mentioned in above should be placed in here."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data reader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import gzip\n",
    "import pickle\n",
    "from os import path\n",
    "from collections import defaultdict\n",
    "from numpy import sign\n",
    "\n",
    "\n",
    "\"\"\"\n",
    "Load buzz data as a dictionary.\n",
    "You can give parameter for data so that you will get what you need only.\n",
    "\"\"\"\n",
    "def load_buzz(root='../data', data=['train', 'test', 'questions'], format='pklz'):\n",
    "    buzz_data = {}\n",
    "    for ii in data:\n",
    "        file_path = path.join(root, ii + \".\" + format)\n",
    "        with gzip.open(file_path, \"rb\") as fp:\n",
    "          buzz_data[ii] = pickle.load(fp)\n",
    "        \n",
    "    return buzz_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature functions(private)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from numpy import sign, abs\n",
    "\n",
    "\n",
    "def _feat_basic(bd, group):\n",
    "    X = []\n",
    "    for item in bd[group].items():\n",
    "        qid = item[1]['qid']\n",
    "        q = bd['questions'][qid]\n",
    "        #item[1]['q_length'] = max(q['pos_token'].keys())\n",
    "        item[1]['q_length'] = len(q['question'].split())\n",
    "        item[1]['category'] = q['category'].lower()\n",
    "        item[1]['answer'] = q['answer'].lower()\n",
    "        X.append(item[1])\n",
    "        \n",
    "    return X\n",
    "        \n",
    "        \n",
    "def _feat_sign_val(data):\n",
    "    for item in data:\n",
    "        item['sign_val'] = sign(item['position'])\n",
    "\n",
    "def _get_pos(bd, sign_val=None):\n",
    "    # bd is not bd, bd is bd['train']\n",
    "    unwanted_index = []\n",
    "    pos_uid = defaultdict(list)\n",
    "    pos_qid = defaultdict(list)\n",
    "    \n",
    "    for index, key in enumerate(bd):\n",
    "        if sign_val and sign(bd[key]['position']) != sign_val:\n",
    "            unwanted_index.append(index)\n",
    "        else:\n",
    "            pos_uid[bd[key]['uid']].append(bd[key]['position'])\n",
    "            pos_qid[bd[key]['qid']].append(bd[key]['position'])\n",
    "    \n",
    "    return pos_uid, pos_qid, unwanted_index\n",
    "\n",
    "\n",
    "def _get_avg_pos(bd, sign_val=None):\n",
    "    pos_uid, pos_qid, unwanted_index = _get_pos(bd, sign_val)\n",
    "\n",
    "    avg_pos_uid = {}\n",
    "    avg_pos_qid = {}\n",
    "    \n",
    "    if not sign_val:\n",
    "        sign_val = 1\n",
    "\n",
    "    for key in pos_uid:\n",
    "        pos = pos_uid[key]\n",
    "        avg_pos_uid[key] = sign_val * (sum(pos) / len(pos))\n",
    "\n",
    "    for key in pos_qid:\n",
    "        pos = pos_qid[key]\n",
    "        avg_pos_qid[key] = sign_val * (sum(pos) / len(pos))\n",
    "    \n",
    "    return avg_pos_uid, avg_pos_qid, unwanted_index\n",
    "\n",
    "        \n",
    "def _feat_avg_pos(data, bd, group, sign_val):\n",
    "    avg_pos_uid, avg_pos_qid, unwanted_index = _get_avg_pos(bd['train'], sign_val=sign_val)\n",
    "    \n",
    "    if group == 'train':\n",
    "        for index in sorted(unwanted_index, reverse=True):\n",
    "            del data[index]\n",
    "    \n",
    "    for item in data:\n",
    "        if item['uid'] in avg_pos_uid:\n",
    "            item['avg_pos_uid'] = avg_pos_uid[item['uid']]\n",
    "        else:\n",
    "            vals = avg_pos_uid.values()\n",
    "            item['avg_pos_uid'] = sum(vals) / float(len(vals))\n",
    "              \n",
    "        if item['qid'] in avg_pos_qid:\n",
    "            item['avg_pos_qid'] = avg_pos_qid[item['qid']]\n",
    "        else:\n",
    "            vals = avg_pos_qid.values()\n",
    "            item['avg_pos_qid'] = sum(vals) / float(len(vals))\n",
    "        \n",
    "        # Response position can be longer than length of question\n",
    "        if item['avg_pos_uid'] > item['q_length']:\n",
    "            item['avg_pos_uid'] = item['q_length']\n",
    "        \n",
    "        if item['avg_pos_qid'] > item['q_length']:\n",
    "            item['avg_pos_qid'] = item['q_length']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature function(public)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def featurize(bd, group, sign_val=None, extra=None):\n",
    "    # Basic features\n",
    "    # qid(string), uid(string), position(float)\n",
    "    # answer'(string), 'potistion'(float), 'qid'(string), 'uid'(string)\n",
    "    X = _feat_basic(bd, group=group)\n",
    "    \n",
    "    # Some extra features\n",
    "    if extra:\n",
    "        for func_name in extra:\n",
    "            func_name = '_feat_' + func_name\n",
    "            if func_name in ['_feat_avg_pos']:\n",
    "                globals()[func_name](X, bd, group=group, sign_val=sign_val)\n",
    "            else:\n",
    "                globals()[func_name](X)\n",
    "    \n",
    "    if group == 'train':\n",
    "        y = []\n",
    "        for item in X:\n",
    "            y.append(item['position'])\n",
    "            del item['position']\n",
    "\n",
    "        return X, y\n",
    "    elif group == 'test':\n",
    "        return X\n",
    "    else:\n",
    "        raise ValueError(group, 'is not the proper type')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Utility functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import csv\n",
    "\n",
    "\n",
    "def select(data, keys):\n",
    "    unwanted = data[0].keys() - keys\n",
    "    for item in data:\n",
    "        for unwanted_key in unwanted:\n",
    "            del item[unwanted_key]\n",
    "    return data\n",
    "\n",
    "\n",
    "def write_result(test_set, predictions, file_name='guess.csv'):\n",
    "    predictions = sorted([[id, predictions[index]] for index, id in enumerate(test_set.keys())])\n",
    "    predictions.insert(0,[\"id\", \"position\"])\n",
    "    with open(file_name, \"w\") as fp:\n",
    "        writer = csv.writer(fp, delimiter=',')\n",
    "        writer.writerows(predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## GMM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Classifying questions\n",
    "features: avg_pos, accuracy rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "pos_uid, pos_qid, _ = _get_pos(load_buzz()['train'], sign_val=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "for key in pos_uid:\n",
    "    q = np.array(pos_uid[key])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "regression_keys = ['category', 'q_length']\n",
    "X_train, y_train = featurize(load_buzz(), group='train', sign_val=None, extra=['sign_val', 'avg_pos'])\n",
    "X_train = select(X_train, regression_keys)\n",
    "X = np.array([[x['q_length'], y_train[index]] for index, x in enumerate(X_train)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "regression_keys = ['category', 'q_length']\n",
    "X_train, y_train = featurize(load_buzz(), group='train', sign_val=None, extra=['sign_val', 'avg_pos'])\n",
    "X_train = select(X_train, regression_keys)\n",
    "X = np.array([[x['q_length'], y_train[index]] for index, x in enumerate(X_train)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjwAAAR5CAYAAADJSt9/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1290abb780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import itertools\n",
    "\n",
    "import numpy as np\n",
    "from scipy import linalg\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "\n",
    "from sklearn import mixture\n",
    "\n",
    "# Fit a mixture of Gaussians with EM using five components\n",
    "gmm = mixture.GMM(n_components=5, covariance_type='full')\n",
    "gmm.fit(X)\n",
    "\n",
    "# Fit a Variational Inference mixture of Gaussians using five components\n",
    "vbgmm = mixture.VBGMM(n_components=5, covariance_type='full')\n",
    "vbgmm.fit(X)\n",
    "\n",
    "# Fit a Dirichlet process mixture of Gaussians using five components\n",
    "dpgmm = mixture.DPGMM(n_components=5, covariance_type='full')\n",
    "dpgmm.fit(X)\n",
    "\n",
    "color_iter = itertools.cycle(['r', 'g', 'b', 'c', 'm'])\n",
    "plt.figure(figsize=(10, 20))\n",
    "for i, (clf, title) in enumerate([(gmm, 'GMM'),\n",
    "                                  (vbgmm, 'Variational Inference for the Gaussian Mixture Model'),\n",
    "                                  (dpgmm, 'Dirichlet Process GMM'),\n",
    "                                 ]):\n",
    "    splot = plt.subplot(3, 1, 1 + i)\n",
    "\n",
    "    Y_ = clf.predict(X)\n",
    "    for i, (mean, covar, color) in enumerate(zip(\n",
    "            clf.means_, clf._get_covars(), color_iter)):\n",
    "        v, w = linalg.eigh(covar)\n",
    "        u = w[0] / linalg.norm(w[0])\n",
    "        # as the DP will not use every component it has access to\n",
    "        # unless it needs it, we shouldn't plot the redundant\n",
    "        # components.\n",
    "        if not np.any(Y_ == i):\n",
    "            continue\n",
    "        plt.scatter(X[Y_ == i, 0], X[Y_ == i, 1], .8, color=color)\n",
    "        \n",
    "        \"\"\"\n",
    "        # Plot an ellipse to show the Gaussian component\n",
    "        angle = np.arctan(u[1] / u[0])\n",
    "        angle = 180 * angle / np.pi  # convert to degrees\n",
    "        ell = mpl.patches.Ellipse(mean, v[0], v[1], 180 + angle, color=color)\n",
    "        ell.set_clip_box(splot.bbox)\n",
    "        ell.set_alpha(0.5)\n",
    "        splot.add_artist(ell)\n",
    "        \"\"\"\n",
    "\n",
    "    plt.xticks(())\n",
    "    plt.yticks(())\n",
    "    plt.title(title)\n",
    "    plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## B. Modeling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Select model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "regression_keys = ['category', 'q_length', 'qid', 'uid', 'answer', 'avg_pos_qid', 'avg_pos_pid']\n",
    "X_train, y_train = featurize(load_buzz(), group='train', sign_val=None, extra=['sign_val', 'avg_pos'])\n",
    "X_train = select(X_train, regression_keys)\n",
    "\n",
    "vec = DictVectorizer()\n",
    "X_train = vec.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== Linear Cross validation RMSE scores:\n",
      "LinearRegression 71.0653342096\n",
      "Ridge 71.4756234707\n",
      "Lasso 71.0311904587\n",
      "ElasticNet 71.0311996227\n"
     ]
    }
   ],
   "source": [
    "import multiprocessing\n",
    "from sklearn import linear_model\n",
    "from sklearn.cross_validation import train_test_split, cross_val_score\n",
    "from sklearn.feature_extraction import DictVectorizer\n",
    "import math\n",
    "from numpy import abs, sqrt\n",
    "\n",
    "\n",
    "regressor_names = \"\"\"\n",
    "LinearRegression\n",
    "Ridge\n",
    "Lasso\n",
    "ElasticNet\n",
    "\"\"\"\n",
    "print (\"=== Linear Cross validation RMSE scores:\")\n",
    "for regressor in regressor_names.split():\n",
    "    scores = cross_val_score(getattr(linear_model, regressor)(),\n",
    "                             X_train, y_train,\n",
    "                             cv=10,\n",
    "                             scoring='mean_squared_error',\n",
    "                             n_jobs=multiprocessing.cpu_count()-1\n",
    "                            )\n",
    "    print (regressor, sqrt(abs(scores)).mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training and testing model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "regression_keys = ['category', 'q_length', 'qid', 'uid', 'answer', 'avg_pos_qid', 'avg_pos_pid']\n",
    "X_train, y_train = featurize(load_buzz(), group='train', sign_val=None, extra=['avg_pos'])\n",
    "X_train = select(X_train, regression_keys)\n",
    "X_test = featurize(load_buzz(), group='test', sign_val=None, extra=['avg_pos'])\n",
    "X_test = select(X_test, regression_keys)\n",
    "\n",
    "vec = DictVectorizer()\n",
    "vec.fit(X_train + X_test)\n",
    "X_train = vec.transform(X_train)\n",
    "X_test = vec.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LassoCV(alphas=None, copy_X=True, cv=None, eps=0.001, fit_intercept=True,\n",
       "    max_iter=1000, n_alphas=100, n_jobs=1, normalize=False, positive=False,\n",
       "    precompute='auto', random_state=None, selection='cyclic', tol=0.0001,\n",
       "    verbose=False)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regressor = linear_model.LassoCV()\n",
    "regressor.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[  0.00000000e+00   0.00000000e+00  -0.00000000e+00 ...,   0.00000000e+00\n",
      "   4.28043156e-05  -1.87619270e-02]\n",
      "146.866398809\n"
     ]
    }
   ],
   "source": [
    "print(regressor.coef_)\n",
    "print(regressor.alpha_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "predictions = regressor.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Writing result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "write_result(bd['test'], predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This submissions scores 84.32000 in Kaggle."
   ]
  }
 ],
 "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": "3.4.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}