finding_song_draft-debug.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Computer: lookfar.local\n"
     ]
    }
   ],
   "source": [
    "#notebook for testing the search of song in a wav file\n",
    "\n",
    "## Load everything\n",
    "%matplotlib inline\n",
    "import pdb\n",
    "import sys\n",
    "import numpy as np\n",
    "import scipy.io as sio\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import scipy.signal as sg\n",
    "import math\n",
    "import scipy as sp\n",
    "import socket\n",
    "import os\n",
    "\n",
    "matplotlib.style.use('ggplot')\n",
    "\n",
    "# Set up Computer and paths\n",
    "comp_name=socket.gethostname()\n",
    "print 'Computer: ' + comp_name\n",
    "if  comp_name == 'chim':\n",
    "    sys.path.append('C:\\Users\\GentnerLab\\Documents\\Experiment\\scripts\\sound_tools')\n",
    "    experiment_folder = os.path.join('C:\\Users\\GentnerLab\\Documents\\Experiment')\n",
    "    r = os.path.join('C:\\Users\\GentnerLab\\Documents\\Experiment')\n",
    "    raw_data_folder = os.path.join(experiment_folder, 'raw_data')\n",
    "\n",
    "if 'lintu' in comp_name:\n",
    "    sys.path.append('/mnt/cube/earneodo/repos/soundflow/sound_tools')\n",
    "    experiment_folder = os.path.join('/mnt/cube/earneodo/bci_zf/')\n",
    "    raw_data_folder = os.path.join(experiment_folder, 'raw_data')\n",
    "\n",
    "if 'lookfar' in comp_name:\n",
    "    sys.path.append('/Users/zeke/repos/soundflow/sound_tools')\n",
    "    experiment_folder = os.path.join('/Users/zeke/experiments/bci_zf/')\n",
    "    raw_data_folder = os.path.join(experiment_folder, 'raw_data')\n",
    "\n",
    "#grab soundtools\n",
    "from soundtools import WavData, Chunk\n",
    "    \n",
    "#file structure\n",
    "bird_id = 'z004'\n",
    "sess = 2\n",
    "rec = 'a'\n",
    "run = 1\n",
    "#bout = '2251_869'\n",
    "bout = None\n",
    "\n",
    "#load the bout and the whole sound file for searching\n",
    "raw_file_folder = os.path.join(raw_data_folder, bird_id, str(sess).zfill(3))\n",
    "raw_file_name = rec + '_' + str(run).zfill(2) + '_song.wav'\n",
    "raw_file_path = os.path.join(raw_file_folder, raw_file_name)\n",
    "\n",
    "bout_file_name = rec + '_' + str(run).zfill(2) + '_autobouts.wav'\n",
    "bout_file_path = os.path.join(raw_file_folder, bout_file_name)\n",
    "\n",
    "#the whole sound\n",
    "all_sound = WavData(raw_file_path)\n",
    "\n",
    "#the bout\n",
    "if bout is not None:\n",
    "    raw_file_folder = os.path.join(raw_file_folder, 'bouts')\n",
    "    raw_file_name = rec + '_' + str(run).zfill(2) + '_song_' + bout + '.wav'\n",
    "\n",
    "raw_file_path = os.path.join(raw_file_folder, raw_file_name)\n",
    "bout_sound = WavData(raw_file_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#functions that should be in a package\n",
    "def rms(x):\n",
    "    return np.linalg.norm(x)\n",
    "\n",
    "def mad(x):\n",
    "    med = np.median(x)\n",
    "    dev = np.abs(x - np.median(x))\n",
    "    return np.median(dev)\n",
    "\n",
    "def identity(x):\n",
    "    return x\n",
    "\n",
    "#decorator for getting the rms after applying a filter to a segment\n",
    "def rms_after_filter(filter_func):\n",
    "    def out_of_filtered(*args, **kwargs):\n",
    "        #print \"Arguments were: %s, %s\" % (args, kwargs)\n",
    "        #plt.plot(np.transpose(filter_func(*args, **kwargs)))\n",
    "        return np.linalg.norm(filter_func(*args, **kwargs))\n",
    "    return out_of_filtered"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# preparations and parameters\n",
    "# filter parameters and filter function\n",
    "#filter the signal\n",
    "s_f = all_sound.s_f\n",
    "\n",
    "#The band where we expect most of the energy of zf song to be in\n",
    "filt_lo = 10000 #Hz\n",
    "filt_hi = 500 #Hz\n",
    "hp_b, hp_a = sg.butter(4, filt_hi/(s_f/2.), btype='high')\n",
    "lp_b, lp_a = sg.butter(4, filt_lo/(s_f/2.), btype='low')\n",
    "\n",
    "\n",
    "def band_pass_filter(chunk, hp_b, hp_a, lp_b, lp_a):    \n",
    "    chunk_hi = sg.filtfilt(hp_b, hp_a, chunk)\n",
    "    chunk_filt = sg.filtfilt(lp_b, lp_a, chunk_hi)\n",
    "    return chunk_filt\n",
    "\n",
    "@rms_after_filter\n",
    "def rms_band_pass_filter(chunk, hp_b, hp_a, lp_b, lp_a):\n",
    "    #plt.plot(np.transpose(chunk))\n",
    "    chunk_hi = sg.filtfilt(hp_b, hp_a, chunk)\n",
    "    chunk_filt = sg.filtfilt(lp_b, lp_a, chunk_hi)\n",
    "    return chunk_filt\n",
    "\n",
    "@rms_after_filter\n",
    "def rms_identity(chunk):\n",
    "    return(chunk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "557.0\n",
      "26.0\n",
      "3157.0\n"
     ]
    }
   ],
   "source": [
    "#Parameters of the search\n",
    "bit_size = 100 # ms; size of sound bits\n",
    "refractory = 5 # bits; refractory period in bits\n",
    "bits_bout = 4 # bits; how many bouts together is considered a bout\n",
    "threshold = 100 # threshold in mads (median deviation of the median)\n",
    "\n",
    "bit_size_sample = np.int(bit_size*all_sound.s_f*0.001)\n",
    "all_starts = np.arange(0, all_sound.n_samples - bit_size_sample, bit_size_sample)\n",
    "\n",
    "all_rms = all_sound.apply_repeated(all_starts, bit_size_sample, rms_band_pass_filter, hp_b, hp_a, lp_b, lp_a)\n",
    "\n",
    "thresh_rms = np.median(all_rms) + threshold * mad(all_rms)\n",
    "\n",
    "print np.median(all_rms)\n",
    "print mad(all_rms)\n",
    "print thresh_rms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1289601d0>]"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEECAYAAADpigmnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhEkREUIIMWFSRIQQQkzY/w8jW/6ZZQI0aAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11c7ac310>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "supra_bits = np.where(all_rms > thresh_rms)[0]\n",
    "plt.plot(np.arange(supra_bits.size), supra_bits, '*')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Starts of probable bouts\n",
    "possible_bouts = np.where(np.diff(supra_bits, bits_bout)==0)[0]\n",
    "\n",
    "#get begginings of bits_bout chunks above threshold\n",
    "firsts = possible_bouts[np.where(np.diff(possible_bouts,1)>1)[0]+1]\n",
    "firsts = np.insert(firsts, 0, possible_bouts[0])\n",
    "\n",
    "lasts = possible_bouts[np.where(np.diff(possible_bouts,1)>1)[0]] + bits_bout\n",
    "lasts = np.insert(lasts, lasts.size, possible_bouts[-1]) \n",
    "\n",
    "#gather segments that are less than refractory apart\n",
    "together = np.where((firsts[1:] - lasts[:-1]) < refractory)[0]\n",
    "\n",
    "lasts = np.delete(lasts, together)\n",
    "firsts = np.delete(firsts, together +1)\n",
    "assert(firsts.size == lasts.size)\n",
    "\n",
    "segments = np.array([all_starts[supra_bits[firsts]], all_starts[supra_bits[lasts]]])\n",
    "segment = np.delete(segments,np.where(np.diff(segments, axis=0)[0]==0)[0], axis=1)\n",
    "assert(np.all(np.diff(segments, axis=0)[0]>0))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "132"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "segment.size\n",
    "segment = np.delete(segments,np.where(np.diff(segments, axis=0)[0]==0)[0], axis=1)\n",
    "segment.size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Another way of finding the bouts:\n",
    "# No refractory period for now\n",
    "\n",
    "from itertools import groupby\n",
    "from operator import itemgetter\n",
    "group_bound = np.array([0, 0], dtype=np.int32)\n",
    "\n",
    "for k, g in groupby(enumerate(supra_bits), lambda (i, x): i-x):\n",
    "    group = map(itemgetter(1), g)\n",
    "    if len(group) >= bits_bout:\n",
    "        group_bound = np.vstack((group_bound, np.array([group[0], group[-1] + 1])) )\n",
    "        #print group\n",
    "\n",
    "group_bound = np.delete(group_bound, 0, axis=0)\n",
    "# group_bound is in index of supra_bits\n",
    "# segments has to be in samples \n",
    "segments = np.array([all_starts[group_bound[:,0]], all_starts[group_bound[:,1]]])\n",
    "assert(np.all(np.diff(segments, axis=0)[0]>0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 15435,  15435,  15435,  15435,  15435,  15435,  15435,  15435,\n",
       "         18522,  21609,  15435,  15435,  18522,  15435,  15435,  15435,\n",
       "         15435,  15435,  15435,  33957,  15435,  15435,  24696,  27783,\n",
       "         15435,  15435,  15435,  15435,  15435,  15435,  15435,  15435,\n",
       "         24696,  27783,  15435,  15435,  15435,  15435,  15435,  15435,\n",
       "         15435,  15435,  15435,  15435,  18522,  18522,  15435,  15435,\n",
       "         15435,  15435,  15435,  15435,  15435,  15435,  15435,  24696,\n",
       "         15435,  18522,  15435,  15435,  15435,  15435,  15435,  18522,\n",
       "         15435,  24696,  15435,  15435,  15435,  15435,  21609,  18522,\n",
       "         24696,  21609,  15435,  18522,  18522,  15435,  24696,  24696,\n",
       "         15435,  15435,  15435,  15435,  24696,  15435,  15435,  15435,\n",
       "         15435,  15435,  15435,  18522,  15435,  21609,  18522,  18522,\n",
       "         18522,  15435,  24696,  15435,  15435,  15435,  18522,  15435,\n",
       "         15435,  15435,  15435,  15435,  15435,  18522,  18522,  15435,\n",
       "         15435,  15435,  18522,  15435,  15435,  18522,  15435,  18522,\n",
       "         15435,  15435,  18522,  15435,  21609,  15435,  18522,  24696,\n",
       "         18522,  15435,  15435,  15435,  18522,  15435,  24696,  18522,\n",
       "         15435,  15435,  15435,  24696,  27783,  27783,  15435,  21609,\n",
       "         15435,  15435,  15435,  15435,  18522,  15435,  15435,  74088,\n",
       "        120393,  52479, 222264,  49392,  40131,  27783,  24696,  18522,\n",
       "         27783,  27783,  74088,  95697,  40131,  24696,  18522,  15435,\n",
       "         24696, 142002,  77175]])"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.diff(segments, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2, 203)"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "segments.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/zeke/experiments/bci_zf/raw_data/z004/002/a_01_autobouts.wav\n"
     ]
    }
   ],
   "source": [
    "#Make a chunk with each segment and record it to a file\n",
    "more = 2 #how many bits before and after to save\n",
    "n_seg = segments.shape[1]\n",
    "\n",
    "# bout_file_name = rec + '_' + str(run).zfill(2) + 'bouts.wav'\n",
    "# bout_file_path = os.path.join(raw_file_folder, bout_file_name)\n",
    "\n",
    "# bout_file_path = '/Users/zeke/bout_test2.wav'\n",
    "\n",
    "import wave\n",
    "import struct\n",
    "\n",
    "bouts_out = wave.open(bout_file_path, 'w')\n",
    "print bout_file_path\n",
    "bouts_out.setparams((all_sound.n_chans, 2, all_sound.s_f, 0, 'NONE', 'not compressed'))\n",
    "\n",
    "for seg in segments.T:\n",
    "    seg = seg + np.array([-more, more])* bit_size_sample\n",
    "\n",
    "    start = seg[0] if seg[0] > 0 else 0\n",
    "    end = seg[1] if seg[1] < all_sound.n_samples else all_sound.n_samples\n",
    "    #print start, end\n",
    "    sg_chunk = Chunk(all_sound, segment = [start, end])\n",
    "    # save the chunk to a data file\n",
    "    #filter for debugging\n",
    "    #sg_chunk.apply_filter(band_pass_filter, hp_b, hp_a, lp_b, lp_a)\n",
    "    data_out = np.array(sg_chunk.data, dtype=np.int16)[0]\n",
    "    packed_data = struct.pack('<' + str((end - start)) + 'h', *data_out)\n",
    "    bouts_out.writeframes(data_out)\n",
    "\n",
    "bouts_out.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZ0AAAEECAYAAADztukpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11c7ac4d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "j = 50\n",
    "more=0\n",
    "chu = Chunk(all_sound, segment=[segments[0, j] - more *bit_size_sample, segments[1, j] + more *bit_size_sample])\n",
    "chu.apply_filter(band_pass_filter, hp_b, hp_a, lp_b, lp_a)\n",
    "\n",
    "chu_starts = np.arange(segments[0, j] - more*bit_size_sample, segments[1,j] + more *bit_size_sample, bit_size_sample)\n",
    "rms_chu = all_sound.apply_repeated(chu_starts, bit_size_sample, rms_band_pass_filter, hp_b, hp_a, lp_b, lp_a)\n",
    "plt.plot(np.transpose(chu.data))\n",
    "plt.plot(chu_starts-chu_starts[0], rms_chu, '*')\n",
    "\n",
    "#just one bit for visualizing and debugging more\n",
    "k = 0\n",
    "test_bit = Chunk(all_sound, segment=[segments[0, j] + k* bit_size_sample, segments[0, j] + (k+1)* bit_size_sample])\n",
    "test_bit.apply_filter(band_pass_filter, hp_b, hp_a, lp_b, lp_a)\n",
    "plt.plot(np.transpose(test_bit.data))\n",
    "rms_test = rms(test_bit.data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 13118, 133973,  14847,   6868])"
      ]
     },
     "execution_count": 259,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rms_chu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([110992, 234015, 186633, 119681,  18371])"
      ]
     },
     "execution_count": 236,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rms_chu[rms_chu>thresh_rms]\n",
    "\n",
    "#print rms_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False, False,  True,  True,  True,  True,  True, False], dtype=bool)"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rms_chu > thresh_rms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "seg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "466.0\n",
      "23.0\n",
      "696.0\n"
     ]
    }
   ],
   "source": [
    "refractory = 2 # bits; refractory period in bits\n",
    "bits_bout = 7 # bits; how many bouts together is considered a bout\n",
    "threshold = 10 # threshold in mads (median deviation of the median)\n",
    "\n",
    "bit_size_sample = np.int(bit_size*all_sound.s_f*0.001)\n",
    "all_starts = np.arange(0, all_sound.n_samples - bit_size_sample, bit_size_sample)\n",
    "\n",
    "all_rms = all_sound.apply_repeated(all_starts, bit_size_sample, rms_band_pass_filter, hp_b, hp_a, lp_b, lp_a)\n",
    "\n",
    "thresh_rms = np.median(all_rms) + threshold * mad(all_rms)\n",
    "\n",
    "print np.median(all_rms)\n",
    "print mad(all_rms)\n",
    "print thresh_rms\n",
    "supra_bits = np.where(all_rms > thresh_rms)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([    0,  3087,  6174,  9261, 12348, 15435, 18522, 21609, 24696,\n",
       "       27783, 30870, 33957, 37044, 40131, 43218, 46305, 49392, 52479,\n",
       "       55566, 58653])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_starts[0:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 110,  191,  227,  244,  245,  297,  320,  401,  416,  447,  571,\n",
       "         691,  754,  755,  756,  757,  758,  759,  825,  826,  827,  835,\n",
       "         836,  837,  838,  985,  986,  987,  993,  994,  995,  996, 1116,\n",
       "        1162, 1163, 1463, 1494, 1500, 1503, 1539, 1583, 1729, 1839, 1840,\n",
       "        1841, 1860, 1922, 2077, 2225, 2226, 2227, 2444, 2445, 2459, 2469,\n",
       "        2470, 2471, 2492, 2493, 2564, 2584, 2644, 2645, 2646, 2660, 2661,\n",
       "        2662, 2776, 2777, 2778, 2831, 2851, 2917, 2938, 2942, 2943, 2951,\n",
       "        2967, 2973, 3028, 3094, 3095, 3096, 3173, 3298, 3408, 3427, 3485,\n",
       "        3523, 3528, 3573, 3647, 3672, 3788, 3809, 3821, 3822, 3871, 3877,\n",
       "        3885, 3886, 3887, 3896, 4034, 4142, 4143, 4144, 4190, 4249, 4250,\n",
       "        4251, 4257, 4258, 4259, 4260, 4266, 4267, 4268, 4269, 4288, 4289,\n",
       "        4484, 4526, 4527, 4528, 4529, 4530, 4531, 4532, 4533, 4534, 4535,\n",
       "        4536, 4537, 4538, 4539, 4540, 4541, 4542, 4543, 4544, 4550, 4551,\n",
       "        4552, 4553, 4554, 4555, 4556, 4557, 4558, 4559, 4560, 4561, 4562,\n",
       "        4563, 4564, 4565, 4566, 4567, 4568, 4569, 4570, 4571, 4572, 4573,\n",
       "        4574, 4575, 4576, 4577, 4578, 4579, 4580, 4581, 4582, 4583, 4589,\n",
       "        4590, 4591, 4592, 4593, 4594, 4595, 4596, 4597, 4598, 4599, 4600,\n",
       "        4606, 4607, 4608, 4609, 4610, 4611, 4612, 4613, 4614, 4615, 4616,\n",
       "        4617, 4618, 4619, 4620, 4621, 4622, 4623, 4624, 4625, 4626, 4627,\n",
       "        4628, 4629, 4630, 4631, 4632, 4633, 4634, 4635, 4636, 4637, 4638,\n",
       "        4639, 4640, 4641, 4642, 4643, 4644, 4645, 4646, 4647, 4648, 4649,\n",
       "        4650, 4651, 4652, 4653, 4654, 4655, 4656, 4657, 4658, 4659, 4660,\n",
       "        4661, 4662, 4663, 4664, 4665, 4666, 4667, 4668, 4669, 4670, 4671,\n",
       "        4672, 4678, 4679, 4680, 4681, 4682, 4683, 4684, 4685, 4686, 4687,\n",
       "        4688, 4696, 4697, 4698, 4699, 4700, 4701, 4702, 4703, 4709, 4710,\n",
       "        4711, 4712, 4718, 4719, 4720, 4726, 4736, 4737, 4738, 4739, 4745,\n",
       "        4746, 4747, 4748, 4754, 4755, 4756, 4757, 4758, 4759, 4760, 4761,\n",
       "        4762, 4763, 4764, 4765, 4766, 4767, 4768, 4769, 4770, 4771, 4772,\n",
       "        4781, 4782, 4783, 4784, 4785, 4786, 4787, 4788, 4789, 4790, 4791,\n",
       "        4792, 4793, 4794, 4795, 4796, 4797, 4798, 4799, 4800, 4801, 4802,\n",
       "        4803, 4804, 4805, 4806, 4812, 4813, 4814, 4815, 4816, 4817, 4818,\n",
       "        4819, 4832, 4833, 4834, 4840, 4863, 4864, 4865, 4882, 4883, 4884,\n",
       "        4885, 4886, 4887, 4888, 4889, 4890, 4891, 4892, 4893, 4894, 4895,\n",
       "        4896, 4897, 4898, 4899, 4900, 4901, 4902, 4903, 4904, 4905, 4906,\n",
       "        4907, 4908, 4909, 4910, 4911, 4912, 4913, 4914, 4915, 4916, 4917,\n",
       "        4918, 4919, 4920, 4921, 4922, 4928, 4929, 4930, 4931, 4932, 4933,\n",
       "        4934, 4935, 4936, 4937, 4938, 4939, 4940, 4941, 4942, 4943, 4944,\n",
       "        4945, 4946, 4947]),)"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pb = np.where(np.diff(supra_bits, 5)==0)\n",
    "pb[0:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([6337, 6341, 6342, 6343, 6347, 6348, 6351, 6372, 6373, 6374, 6375,\n",
       "       6378])"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "supra_bits[108:120]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1599, -1041,   398, ...,    65,   -48,    27])"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.diff(supra_bits, 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[5671, 5672, 5673, 5674, 5675]\n",
      "[5831, 5832, 5833, 5834, 5835]\n",
      "[6036, 6037, 6038, 6039, 6040]\n",
      "[6129, 6130, 6131, 6132, 6133]\n",
      "[6561, 6562, 6563, 6564, 6565]\n",
      "[6607, 6608, 6609, 6610, 6611]\n",
      "[6889, 6890, 6891, 6892, 6893]\n",
      "[6938, 6939, 6940, 6941, 6942]\n",
      "[6960, 6961, 6962, 6963, 6964, 6965]\n",
      "[7011, 7012, 7013, 7014, 7015, 7016, 7017]\n",
      "[7144, 7145, 7146, 7147, 7148]\n",
      "[7230, 7231, 7232, 7233, 7234]\n",
      "[7782, 7783, 7784, 7785, 7786, 7787]\n",
      "[8747, 8748, 8749, 8750, 8751]\n",
      "[8799, 8800, 8801, 8802, 8803]\n",
      "[9113, 9114, 9115, 9116, 9117]\n",
      "[9142, 9143, 9144, 9145, 9146]\n",
      "[9332, 9333, 9334, 9335, 9336]\n",
      "[9716, 9717, 9718, 9719, 9720]\n",
      "[9873, 9874, 9875, 9876, 9877, 9878, 9879, 9880, 9881, 9882, 9883]\n",
      "[9942, 9943, 9944, 9945, 9946]\n",
      "[9982, 9983, 9984, 9985, 9986]\n",
      "[10549, 10550, 10551, 10552, 10553, 10554, 10555, 10556]\n",
      "[10591, 10592, 10593, 10594, 10595, 10596, 10597, 10598, 10599]\n",
      "[10650, 10651, 10652, 10653, 10654]\n",
      "[10660, 10661, 10662, 10663, 10664]\n",
      "[10706, 10707, 10708, 10709, 10710]\n",
      "[10765, 10766, 10767, 10768, 10769]\n",
      "[10792, 10793, 10794, 10795, 10796]\n",
      "[10870, 10871, 10872, 10873, 10874]\n",
      "[11033, 11034, 11035, 11036, 11037]\n",
      "[11474, 11475, 11476, 11477, 11478]\n",
      "[11492, 11493, 11494, 11495, 11496, 11497, 11498, 11499]\n",
      "[11512, 11513, 11514, 11515, 11516, 11517, 11518, 11519, 11520]\n",
      "[11605, 11606, 11607, 11608, 11609]\n",
      "[12589, 12590, 12591, 12592, 12593]\n",
      "[12638, 12639, 12640, 12641, 12642]\n",
      "[12993, 12994, 12995, 12996, 12997]\n",
      "[13147, 13148, 13149, 13150, 13151]\n",
      "[13264, 13265, 13266, 13267, 13268]\n",
      "[13273, 13274, 13275, 13276, 13277]\n",
      "[13352, 13353, 13354, 13355, 13356]\n",
      "[13557, 13558, 13559, 13560, 13561]\n",
      "[13608, 13609, 13610, 13611, 13612]\n",
      "[13828, 13829, 13830, 13831, 13832, 13833]\n",
      "[13936, 13937, 13938, 13939, 13940, 13941]\n",
      "[14003, 14004, 14005, 14006, 14007]\n",
      "[14037, 14038, 14039, 14040, 14041]\n",
      "[14089, 14090, 14091, 14092, 14093]\n",
      "[14364, 14365, 14366, 14367, 14368]\n",
      "[14374, 14375, 14376, 14377, 14378]\n",
      "[14512, 14513, 14514, 14515, 14516]\n",
      "[14649, 14650, 14651, 14652, 14653]\n",
      "[14766, 14767, 14768, 14769, 14770]\n",
      "[16267, 16268, 16269, 16270, 16271]\n",
      "[16285, 16286, 16287, 16288, 16289, 16290, 16291, 16292]\n",
      "[16337, 16338, 16339, 16340, 16341]\n",
      "[16382, 16383, 16384, 16385, 16386, 16387]\n",
      "[16527, 16528, 16529, 16530, 16531]\n",
      "[16708, 16709, 16710, 16711, 16712]\n",
      "[16862, 16863, 16864, 16865, 16866]\n",
      "[16930, 16931, 16932, 16933, 16934]\n",
      "[17040, 17041, 17042, 17043, 17044]\n",
      "[17560, 17561, 17562, 17563, 17564, 17565]\n",
      "[18051, 18052, 18053, 18054, 18055]\n",
      "[18277, 18278, 18279, 18280, 18281, 18282, 18283, 18284]\n",
      "[18383, 18384, 18385, 18386, 18387]\n",
      "[19084, 19085, 19086, 19087, 19088]\n",
      "[19099, 19100, 19101, 19102, 19103]\n",
      "[19130, 19131, 19132, 19133, 19134]\n",
      "[19283, 19284, 19285, 19286, 19287, 19288, 19289]\n",
      "[19374, 19375, 19376, 19377, 19378, 19379]\n",
      "[19404, 19405, 19406, 19407, 19408, 19409, 19410, 19411]\n",
      "[19487, 19488, 19489, 19490, 19491, 19492, 19493]\n",
      "[19651, 19652, 19653, 19654, 19655]\n",
      "[19799, 19800, 19801, 19802, 19803, 19804]\n",
      "[19871, 19872, 19873, 19874, 19875, 19876]\n",
      "[20014, 20015, 20016, 20017, 20018]\n",
      "[20675, 20676, 20677, 20678, 20679, 20680, 20681, 20682]\n",
      "[20737, 20738, 20739, 20740, 20741, 20742, 20743, 20744]\n",
      "[20773, 20774, 20775, 20776, 20777]\n",
      "[20793, 20794, 20795, 20796, 20797]\n",
      "[21104, 21105, 21106, 21107, 21108]\n",
      "[21259, 21260, 21261, 21262, 21263]\n",
      "[21400, 21401, 21402, 21403, 21404, 21405, 21406, 21407]\n",
      "[21563, 21564, 21565, 21566, 21567]\n",
      "[21627, 21628, 21629, 21630, 21631]\n",
      "[21666, 21667, 21668, 21669, 21670]\n",
      "[21746, 21747, 21748, 21749, 21750]\n",
      "[21785, 21786, 21787, 21788, 21789]\n",
      "[21945, 21946, 21947, 21948, 21949]\n",
      "[21971, 21972, 21973, 21974, 21975, 21976]\n",
      "[22007, 22008, 22009, 22010, 22011]\n",
      "[22057, 22058, 22059, 22060, 22061, 22062, 22063]\n",
      "[22078, 22079, 22080, 22081, 22082, 22083]\n",
      "[22151, 22152, 22153, 22154, 22155, 22156]\n",
      "[22170, 22171, 22172, 22173, 22174, 22175]\n",
      "[22455, 22456, 22457, 22458, 22459]\n",
      "[22857, 22858, 22859, 22860, 22861, 22862, 22863, 22864]\n",
      "[23080, 23081, 23082, 23083, 23084]\n",
      "[23563, 23564, 23565, 23566, 23567]\n",
      "[23626, 23627, 23628, 23629, 23630]\n",
      "[23838, 23839, 23840, 23841, 23842, 23843]\n",
      "[23874, 23875, 23876, 23877, 23878]\n",
      "[24162, 24163, 24164, 24165, 24166]\n",
      "[24273, 24274, 24275, 24276, 24277]\n",
      "[24431, 24432, 24433, 24434, 24435]\n",
      "[24911, 24912, 24913, 24914, 24915]\n",
      "[24949, 24950, 24951, 24952, 24953]\n",
      "[24960, 24961, 24962, 24963, 24964, 24965]\n",
      "[25040, 25041, 25042, 25043, 25044, 25045]\n",
      "[25049, 25050, 25051, 25052, 25053]\n",
      "[25084, 25085, 25086, 25087, 25088]\n",
      "[25148, 25149, 25150, 25151, 25152]\n",
      "[25228, 25229, 25230, 25231, 25232, 25233]\n",
      "[25249, 25250, 25251, 25252, 25253]\n",
      "[25280, 25281, 25282, 25283, 25284]\n",
      "[25363, 25364, 25365, 25366, 25367, 25368]\n",
      "[25463, 25464, 25465, 25466, 25467]\n",
      "[25987, 25988, 25989, 25990, 25991, 25992]\n",
      "[26177, 26178, 26179, 26180, 26181]\n",
      "[26535, 26536, 26537, 26538, 26539]\n",
      "[26636, 26637, 26638, 26639, 26640, 26641]\n",
      "[26722, 26723, 26724, 26725, 26726]\n",
      "[26757, 26758, 26759, 26760, 26761, 26762, 26763]\n",
      "[26907, 26908, 26909, 26910, 26911]\n",
      "[27021, 27022, 27023, 27024, 27025, 27026]\n",
      "[27041, 27042, 27043, 27044, 27045, 27046, 27047, 27048]\n",
      "[27073, 27074, 27075, 27076, 27077, 27078]\n",
      "[27141, 27142, 27143, 27144, 27145]\n",
      "[27698, 27699, 27700, 27701, 27702]\n",
      "[27720, 27721, 27722, 27723, 27724]\n",
      "[27786, 27787, 27788, 27789, 27790, 27791]\n",
      "[28074, 28075, 28076, 28077, 28078]\n",
      "[28327, 28328, 28329, 28330, 28331, 28332, 28333, 28334]\n",
      "[28540, 28541, 28542, 28543, 28544, 28545]\n",
      "[28588, 28589, 28590, 28591, 28592]\n",
      "[28701, 28702, 28703, 28704, 28705]\n",
      "[29177, 29178, 29179, 29180, 29181]\n",
      "[29724, 29725, 29726, 29727, 29728, 29729, 29730, 29731]\n",
      "[29738, 29739, 29740, 29741, 29742, 29743, 29744, 29745, 29746]\n",
      "[29755, 29756, 29757, 29758, 29759, 29760, 29761, 29762, 29763]\n",
      "[29800, 29801, 29802, 29803, 29804]\n",
      "[29842, 29843, 29844, 29845, 29846, 29847, 29848]\n",
      "[34054, 34055, 34056, 34057, 34058]\n",
      "[35435, 35436, 35437, 35438, 35439]\n",
      "[46140, 46141, 46142, 46143, 46144]\n",
      "[46195, 46196, 46197, 46198, 46199]\n",
      "[50400, 50401, 50402, 50403, 50404, 50405]\n",
      "[50548, 50549, 50550, 50551, 50552]\n",
      "[54403, 54404, 54405, 54406, 54407]\n",
      "[54409, 54410, 54411, 54412, 54413, 54414, 54415, 54416, 54417, 54418, 54419, 54420, 54421, 54422, 54423, 54424, 54425, 54426, 54427, 54428, 54429, 54430, 54431, 54432]\n",
      "[54434, 54435, 54436, 54437, 54438, 54439, 54440, 54441, 54442, 54443, 54444, 54445, 54446, 54447, 54448, 54449, 54450, 54451, 54452, 54453, 54454, 54455, 54456, 54457, 54458, 54459, 54460, 54461, 54462, 54463, 54464, 54465, 54466, 54467, 54468, 54469, 54470, 54471, 54472]\n",
      "[54474, 54475, 54476, 54477, 54478, 54479, 54480, 54481, 54482, 54483, 54484, 54485, 54486, 54487, 54488, 54489, 54490]\n",
      "[54493, 54494, 54495, 54496, 54497, 54498, 54499, 54500, 54501, 54502, 54503, 54504, 54505, 54506, 54507, 54508, 54509, 54510, 54511, 54512, 54513, 54514, 54515, 54516, 54517, 54518, 54519, 54520, 54521, 54522, 54523, 54524, 54525, 54526, 54527, 54528, 54529, 54530, 54531, 54532, 54533, 54534, 54535, 54536, 54537, 54538, 54539, 54540, 54541, 54542, 54543, 54544, 54545, 54546, 54547, 54548, 54549, 54550, 54551, 54552, 54553, 54554, 54555, 54556, 54557, 54558, 54559, 54560, 54561, 54562, 54563, 54564]\n",
      "[54567, 54568, 54569, 54570, 54571, 54572, 54573, 54574, 54575, 54576, 54577, 54578, 54579, 54580, 54581, 54582]\n",
      "[54587, 54588, 54589, 54590, 54591, 54592, 54593, 54594, 54595, 54596, 54597, 54598, 54599]\n",
      "[54603, 54604, 54605, 54606, 54607, 54608, 54609, 54610, 54611]\n",
      "[54615, 54616, 54617, 54618, 54619, 54620, 54621, 54622]\n",
      "[54624, 54625, 54626, 54627, 54628, 54629]\n",
      "[54637, 54638, 54639, 54640, 54641, 54642, 54643, 54644, 54645]\n",
      "[54648, 54649, 54650, 54651, 54652, 54653, 54654, 54655, 54656]\n",
      "[54659, 54660, 54661, 54662, 54663, 54664, 54665, 54666, 54667, 54668, 54669, 54670, 54671, 54672, 54673, 54674, 54675, 54676, 54677, 54678, 54679, 54680, 54681, 54682]\n",
      "[54689, 54690, 54691, 54692, 54693, 54694, 54695, 54696, 54697, 54698, 54699, 54700, 54701, 54702, 54703, 54704, 54705, 54706, 54707, 54708, 54709, 54710, 54711, 54712, 54713, 54714, 54715, 54716, 54717, 54718, 54719]\n",
      "[54721, 54722, 54723, 54724, 54725, 54726, 54727, 54728, 54729, 54730, 54731, 54732, 54733]\n",
      "[54748, 54749, 54750, 54751, 54752, 54753, 54754, 54755]\n",
      "[54766, 54767, 54768, 54769, 54770, 54771]\n",
      "[54788, 54789, 54790, 54791, 54792]\n",
      "[54823, 54824, 54825, 54826, 54827, 54828, 54829, 54830]\n",
      "[54867, 54868, 54869, 54870, 54871, 54872, 54873, 54874, 54875, 54876, 54877, 54878, 54879, 54880, 54881, 54882, 54883, 54884, 54885, 54886, 54887, 54888, 54889, 54890, 54891, 54892, 54893, 54894, 54895, 54896, 54897, 54898, 54899, 54900, 54901, 54902, 54903, 54904, 54905, 54906, 54907, 54908, 54909, 54910, 54911, 54912]\n",
      "[54914, 54915, 54916, 54917, 54918, 54919, 54920, 54921, 54922, 54923, 54924, 54925, 54926, 54927, 54928, 54929, 54930, 54931, 54932, 54933, 54934, 54935, 54936, 54937, 54938]\n"
     ]
    }
   ],
   "source": [
    "from itertools import groupby\n",
    "from operator import itemgetter\n",
    "group_bound = np.array([0, 0], dtype=np.int32)\n",
    "\n",
    "for k, g in groupby(enumerate(supra_bits), lambda (i, x): i-x):\n",
    "    group = map(itemgetter(1), g)\n",
    "    if len(group) >= bits_bout:\n",
    "        group_bound = np.vstack((group_bound, np.array([group[0], group[-1]])) )\n",
    "        print group\n",
    "group_bound = np.delete(group_bound, 0, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 5671,  5675],\n",
       "       [ 5831,  5835],\n",
       "       [ 6036,  6040],\n",
       "       [ 6129,  6133],\n",
       "       [ 6561,  6565],\n",
       "       [ 6607,  6611],\n",
       "       [ 6889,  6893],\n",
       "       [ 6938,  6942],\n",
       "       [ 6960,  6965],\n",
       "       [ 7011,  7017],\n",
       "       [ 7144,  7148],\n",
       "       [ 7230,  7234],\n",
       "       [ 7782,  7787],\n",
       "       [ 8747,  8751],\n",
       "       [ 8799,  8803],\n",
       "       [ 9113,  9117],\n",
       "       [ 9142,  9146],\n",
       "       [ 9332,  9336],\n",
       "       [ 9716,  9720],\n",
       "       [ 9873,  9883],\n",
       "       [ 9942,  9946],\n",
       "       [ 9982,  9986],\n",
       "       [10549, 10556],\n",
       "       [10591, 10599],\n",
       "       [10650, 10654],\n",
       "       [10660, 10664],\n",
       "       [10706, 10710],\n",
       "       [10765, 10769],\n",
       "       [10792, 10796],\n",
       "       [10870, 10874],\n",
       "       [11033, 11037],\n",
       "       [11474, 11478],\n",
       "       [11492, 11499],\n",
       "       [11512, 11520],\n",
       "       [11605, 11609],\n",
       "       [12589, 12593],\n",
       "       [12638, 12642],\n",
       "       [12993, 12997],\n",
       "       [13147, 13151],\n",
       "       [13264, 13268],\n",
       "       [13273, 13277],\n",
       "       [13352, 13356],\n",
       "       [13557, 13561],\n",
       "       [13608, 13612],\n",
       "       [13828, 13833],\n",
       "       [13936, 13941],\n",
       "       [14003, 14007],\n",
       "       [14037, 14041],\n",
       "       [14089, 14093],\n",
       "       [14364, 14368],\n",
       "       [14374, 14378],\n",
       "       [14512, 14516],\n",
       "       [14649, 14653],\n",
       "       [14766, 14770],\n",
       "       [16267, 16271],\n",
       "       [16285, 16292],\n",
       "       [16337, 16341],\n",
       "       [16382, 16387],\n",
       "       [16527, 16531],\n",
       "       [16708, 16712],\n",
       "       [16862, 16866],\n",
       "       [16930, 16934],\n",
       "       [17040, 17044],\n",
       "       [17560, 17565],\n",
       "       [18051, 18055],\n",
       "       [18277, 18284],\n",
       "       [18383, 18387],\n",
       "       [19084, 19088],\n",
       "       [19099, 19103],\n",
       "       [19130, 19134],\n",
       "       [19283, 19289],\n",
       "       [19374, 19379],\n",
       "       [19404, 19411],\n",
       "       [19487, 19493],\n",
       "       [19651, 19655],\n",
       "       [19799, 19804],\n",
       "       [19871, 19876],\n",
       "       [20014, 20018],\n",
       "       [20675, 20682],\n",
       "       [20737, 20744],\n",
       "       [20773, 20777],\n",
       "       [20793, 20797],\n",
       "       [21104, 21108],\n",
       "       [21259, 21263],\n",
       "       [21400, 21407],\n",
       "       [21563, 21567],\n",
       "       [21627, 21631],\n",
       "       [21666, 21670],\n",
       "       [21746, 21750],\n",
       "       [21785, 21789],\n",
       "       [21945, 21949],\n",
       "       [21971, 21976],\n",
       "       [22007, 22011],\n",
       "       [22057, 22063],\n",
       "       [22078, 22083],\n",
       "       [22151, 22156],\n",
       "       [22170, 22175],\n",
       "       [22455, 22459],\n",
       "       [22857, 22864],\n",
       "       [23080, 23084],\n",
       "       [23563, 23567],\n",
       "       [23626, 23630],\n",
       "       [23838, 23843],\n",
       "       [23874, 23878],\n",
       "       [24162, 24166],\n",
       "       [24273, 24277],\n",
       "       [24431, 24435],\n",
       "       [24911, 24915],\n",
       "       [24949, 24953],\n",
       "       [24960, 24965],\n",
       "       [25040, 25045],\n",
       "       [25049, 25053],\n",
       "       [25084, 25088],\n",
       "       [25148, 25152],\n",
       "       [25228, 25233],\n",
       "       [25249, 25253],\n",
       "       [25280, 25284],\n",
       "       [25363, 25368],\n",
       "       [25463, 25467],\n",
       "       [25987, 25992],\n",
       "       [26177, 26181],\n",
       "       [26535, 26539],\n",
       "       [26636, 26641],\n",
       "       [26722, 26726],\n",
       "       [26757, 26763],\n",
       "       [26907, 26911],\n",
       "       [27021, 27026],\n",
       "       [27041, 27048],\n",
       "       [27073, 27078],\n",
       "       [27141, 27145],\n",
       "       [27698, 27702],\n",
       "       [27720, 27724],\n",
       "       [27786, 27791],\n",
       "       [28074, 28078],\n",
       "       [28327, 28334],\n",
       "       [28540, 28545],\n",
       "       [28588, 28592],\n",
       "       [28701, 28705],\n",
       "       [29177, 29181],\n",
       "       [29724, 29731],\n",
       "       [29738, 29746],\n",
       "       [29755, 29763],\n",
       "       [29800, 29804],\n",
       "       [29842, 29848],\n",
       "       [34054, 34058],\n",
       "       [35435, 35439],\n",
       "       [46140, 46144],\n",
       "       [46195, 46199],\n",
       "       [50400, 50405],\n",
       "       [50548, 50552],\n",
       "       [54403, 54407],\n",
       "       [54409, 54432],\n",
       "       [54434, 54472],\n",
       "       [54474, 54490],\n",
       "       [54493, 54564],\n",
       "       [54567, 54582],\n",
       "       [54587, 54599],\n",
       "       [54603, 54611],\n",
       "       [54615, 54622],\n",
       "       [54624, 54629],\n",
       "       [54637, 54645],\n",
       "       [54648, 54656],\n",
       "       [54659, 54682],\n",
       "       [54689, 54719],\n",
       "       [54721, 54733],\n",
       "       [54748, 54755],\n",
       "       [54766, 54771],\n",
       "       [54788, 54792],\n",
       "       [54823, 54830],\n",
       "       [54867, 54912],\n",
       "       [54914, 54938]])"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "group_bound"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[54969, 54970, 54971]"
      ]
     },
     "execution_count": 239,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "group_bound = np.zeros((1,2), dtype=np.int32)\n",
    "group_bound[:] = np.nan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-2147483648, -2147483648]], dtype=int32)"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "group_bound"
   ]
  },
  {
   "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": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}