example-cache.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hello, Ignite ML with Ignite Cache!\n",
    "\n",
    "This document contains example of Ignite ML Python API for training on Ignite Cache."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import make_regression\n",
    "from sklearn.datasets import make_classification \n",
    "from sklearn.datasets import make_blobs\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Regression\n",
    "\n",
    "Lets prepare a regression dataset using scikit-learn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x, y = make_regression(n_targets=1)\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from ignite_ml.cache import Ignition\n",
    "\n",
    "ignite = Ignition.ignite(\"/home/gridgain/ignite/examples/config/example-ignite.xml\")\n",
    "train_cache = ignite.createCache(\"test-regression\", parts=1)\n",
    "\n",
    "xy_train = np.column_stack((x_train, y_train))\n",
    "\n",
    "for i in range(xy_train.shape[0]):\n",
    "    train_cache.put(i, xy_train[i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1. Linear Regression\n",
    "\n",
    "Now lets train a linear model using Ignite ML and estimate the quality of the model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8077042184689166"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.regression import LinearRegressionTrainer\n",
    "trainer = LinearRegressionTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "r2_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And just to compare lets do the same for scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8060119825951335"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "reg = LinearRegression()\n",
    "reg.fit(x_train, y_train)\n",
    "\n",
    "r2_score(y_test, reg.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.2. Decision Tree Regression\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.07843534550191322"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.regression import DecisionTreeRegressionTrainer\n",
    "trainer = DecisionTreeRegressionTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "r2_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.022489821096110196"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeRegressor\n",
    "reg = DecisionTreeRegressor()\n",
    "reg.fit(x_train, y_train)\n",
    "\n",
    "r2_score(y_test, reg.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.3. KNN Regression\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2320725883906889"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.regression import KNNRegressionTrainer\n",
    "trainer = KNNRegressionTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "r2_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2320725883906889"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "reg = KNeighborsRegressor()\n",
    "reg.fit(x_train, y_train)\n",
    "\n",
    "r2_score(y_test, reg.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.4 Random Forest Regression\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# TODO: Random Forest requires features metadata.\n",
    "\n",
    "#from ignite_ml.regression import RandomForestRegressionTrainer\n",
    "#trainer = RandomForestRegressionTrainer()\n",
    "#model = trainer.fit_on_cache(train_cache)\n",
    "#\n",
    "#r2_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/gridgain/anaconda3/lib/python3.6/site-packages/sklearn/ensemble/weight_boosting.py:29: DeprecationWarning: numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n",
      "  from numpy.core.umath_tests import inner1d\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.304445419142198"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "reg = RandomForestRegressor()\n",
    "reg.fit(x_train, y_train)\n",
    "\n",
    "r2_score(y_test, reg.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.5 MLP Regression\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8088167603095694"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.common import MLPArchitecture\n",
    "from ignite_ml.regression import MLPRegressionTrainer\n",
    "\n",
    "trainer = MLPRegressionTrainer(MLPArchitecture(100).with_layer(1, activator='linear'))\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "r2_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/gridgain/anaconda3/lib/python3.6/site-packages/sklearn/neural_network/multilayer_perceptron.py:564: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (1000) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.3512421502827394"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.neural_network import MLPRegressor\n",
    "reg = MLPRegressor(max_iter=1000)\n",
    "reg.fit(x_train, y_train)\n",
    "\n",
    "r2_score(y_test, reg.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Classification\n",
    "\n",
    "Lets prepare a classification dataset using scikit-learn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x, y = make_classification()\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ignite = Ignition.ignite(\"/home/gridgain/ignite/examples/config/example-ignite.xml\")\n",
    "train_cache = ignite.createCache(\"test-classification\")\n",
    "\n",
    "xy_train = np.column_stack((x_train, y_train))\n",
    "\n",
    "for i in range(xy_train.shape[0]):\n",
    "    train_cache.put(i, xy_train[i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.1 Decision Tree Classification\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.96"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.classification import DecisionTreeClassificationTrainer\n",
    "trainer = DecisionTreeClassificationTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.84"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "clf = DecisionTreeClassifier()\n",
    "clf.fit(x_train, y_train)\n",
    "\n",
    "accuracy_score(y_test, clf.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2 ANN Classification\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.52"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.classification import ANNClassificationTrainer\n",
    "trainer = ANNClassificationTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3 KNN Classification\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.68"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.classification import KNNClassificationTrainer\n",
    "trainer = KNNClassificationTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.68"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "clf = KNeighborsClassifier()\n",
    "clf.fit(x_train, y_train)\n",
    "\n",
    "accuracy_score(y_test, clf.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### 2.4 LogReg Classification\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.72"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.classification import LogRegClassificationTrainer\n",
    "trainer = LogRegClassificationTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.72"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "clf = LogisticRegression()\n",
    "clf.fit(x_train, y_train)\n",
    "\n",
    "accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.5 SVM Classification\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.92"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ignite_ml.classification import SVMClassificationTrainer\n",
    "trainer = SVMClassificationTrainer()\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.92"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import LinearSVC\n",
    "clf = LinearSVC()\n",
    "clf.fit(x_train, y_train)\n",
    "\n",
    "accuracy_score(y_test, clf.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.6 Random Forest Classification\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# TODO: Random Forest requires features metadata.\n",
    "\n",
    "#from ignite_ml.classification import RandomForestClassificationTrainer\n",
    "#trainer = RandomForestClassificationTrainer()\n",
    "# model = trainer.fit_on_cache(train_cache)\n",
    "# \n",
    "# accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.92"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "clf = RandomForestClassifier()\n",
    "clf.fit(x_train, y_train)\n",
    "\n",
    "accuracy_score(y_test, model.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.7 MLP Classification\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from ignite_ml.common import MLPArchitecture\n",
    "from ignite_ml.classification import MLPClassificationTrainer\n",
    "\n",
    "def encode_label(x):\n",
    "    if x:\n",
    "        return [0, 1]\n",
    "    else:\n",
    "        return [1, 0]\n",
    "\n",
    "def decode_label(x):\n",
    "    if x[0] > x[1]:\n",
    "        return 0\n",
    "    else:\n",
    "        return 1\n",
    "    \n",
    "trainer = MLPRegressionTrainer(MLPArchitecture(20).with_layer(2, activator='sigmoid'))\n",
    "# model = trainer.fit_on_cache(x_train, [encode_label(x) for x in y_train])\n",
    "\n",
    "# accuracy_score(y_test, [decode_label(x) for x in model.predict(x_test)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Clustering\n",
    "\n",
    "Lets prepare a clustering dataset using scikit-learn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x, y = make_blobs(n_samples=2000, n_features=2, cluster_std=1.0, centers=[(-3, -3), (0, 0), (3, 3)])\n",
    "\n",
    "train_cache = ignite.createCache(\"test-clustering\")\n",
    "\n",
    "xy = np.column_stack((x, y))\n",
    "\n",
    "for i in range(xy.shape[0]):\n",
    "    train_cache.put(i, xy[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsdglqRiLXZKKsdglqZg3WkOQkjNGuKkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcd480c3940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(3):\n",
    "    plt.scatter(x[y == i][:, 0], x[y == i][:, 1], s=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1 KMeans Clustering\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcd50292c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from ignite_ml.clustering import KMeansClusteringTrainer\n",
    "trainer = KMeansClusteringTrainer(amount_of_clusters=2)\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "y_predicted = np.array(model.predict(x))\n",
    "for i in range(3):\n",
    "    plt.scatter(x[y_predicted == i][:, 0], x[y_predicted == i][:, 1], s=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And using scikit-learn:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcd381ad160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "clr = KMeans(n_clusters=2)\n",
    "clr.fit(x)\n",
    "\n",
    "y_predicted = np.array(clr.predict(x))\n",
    "for i in range(2):\n",
    "    plt.scatter(x[y_predicted == i][:, 0], x[y_predicted == i][:, 1], s=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2 GMM Clustering\n",
    "\n",
    "Using Ignite ML:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+WMOmidiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcd4805f940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from ignite_ml.clustering import GMMClusteringTrainer\n",
    "trainer = GMMClusteringTrainer(count_of_components=2, max_count_of_clusters=2)\n",
    "model = trainer.fit_on_cache(train_cache)\n",
    "\n",
    "y_predicted = np.array(model.predict(x))\n",
    "for i in range(2):\n",
    "    plt.scatter(x[y_predicted == i][:, 0], x[y_predicted == i][:, 1], s=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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
}