{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# chemdiab dataset example\n",
    "\n",
    "Data fom\n",
    "[locfit R package](https://cran.r-project.org/web/packages/locfit/locfit.pdf).\n",
    "\n",
    "Numeric variables are rw, fpg, ga, ina and sspg. Classifier cc is the Diabetic type.\n",
    "\n",
    "Originally from **Reaven, G. M. and Miller, R. G. (1979). An attempt to define the nature of chemical diabetes using a multidimensional analysis. Diabetologia 16, 17-24.**\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# import data\n",
    "import numpy as np\n",
    "import inspect\n",
    "from os import path\n",
    "dtype = [(\"rw\",np.float64),(\"fpg\",np.float64),(\"ga\",np.float64),\n",
    "         (\"ina\",np.float64),('sspg', np.float64),('cc', 'S20')]\n",
    "file_path = path.dirname(path.abspath(\n",
    "                         inspect.getfile(inspect.currentframe())))\n",
    "file_path = path.join(path.dirname(file_path),\"chemdiab.tab\")\n",
    "chemdiab = np.genfromtxt(\"chemdiab.tab\", names=True,\n",
    "                         dtype=dtype)\n",
    "# get only numeric columns as numpy array\n",
    "nam_cols = list(chemdiab.dtype.names[0:5])\n",
    "num_cols = chemdiab[nam_cols]\n",
    "num_cols = num_cols.view(np.float64)\\\n",
    "                   .reshape(num_cols.shape + (-1,))\n",
    "\n",
    "cat, ind = np.unique(chemdiab[chemdiab.dtype.names[5]],\n",
    "                     return_inverse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# clean and visualize data\n",
    "from sklearn.preprocessing import StandardScaler, FunctionTransformer\n",
    "from cartographer.filterers import KernelDensityFilterer\n",
    "from seaborn import plt\n",
    "\n",
    "std_scaler = StandardScaler()\n",
    "scaled = std_scaler.fit_transform(num_cols)\n",
    "\n",
    "fig, axs = plt.subplots(2,3, figsize=(16,8))\n",
    "flat_axs = axs.flatten()\n",
    "for i, name in enumerate(nam_cols):\n",
    "    flat_axs[i].hist(scaled[:,i])\n",
    "    flat_axs[i].set_title(name)\n",
    "    \n",
    "flat_axs[5].hist(ind)\n",
    "flat_axs[5].set_title(\"category\")\n",
    "\n",
    "fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from cartographer.mapper import Mapper\n",
    "from cartographer.coverers import HyperRectangleCoverer\n",
    "from sklearn.cluster import DBSCAN\n",
    "from cartographer.visualization import html_graph\n",
    "from IPython.core.display import HTML\n",
    "\n",
    "m = Mapper(coverer=HyperRectangleCoverer(intervals=10, overlap=0.5),\n",
    "           filterer=KernelDensityFilterer(bandwidth=1.0),\n",
    "           clusterer=DBSCAN(min_samples=5,eps=2.))\n",
    "m.fit(scaled)\n",
    "\n",
    "HTML(html_graph(m, {\"ind\": ind}, {\"kde\": m.filterer.transform(scaled)}))"
   ]
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}