ks2_30kHz_continuous_withProbeFile.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Attempt to run Kilosort2 via spike interfaces and using a probe geometry file\n",
    "The data used for this purpose is on continuously sampled wideband data collected by Jimmie Gmaz. [(Source)](http://datasets.datalad.org/?dir=/labs/mvdm/BiconditionalOdor)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prerequisites\n",
    "- Computer with NVIDIA GPU\n",
    "- Install Matlab and compatible CUDA libraries\n",
    "- Download Kilosort2 and compile all the MEX files in the 'CUDA folder'. For details follow the documentation on the [Kilosort2 repository.](https://github.com/MouseLand/Kilosort2/)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Recommended import aliases: https://spikeinterface.readthedocs.io/en/latest/getting_started/plot_getting_started.html\n",
    "import spikeinterface.extractors as se\n",
    "import spikeinterface.toolkit as st\n",
    "import spikeinterface.sorters as ss\n",
    "import spikeinterface.comparison as sc\n",
    "import spikeinterface.widgets as sw\n",
    "# Additional imports\n",
    "import os.path as op\n",
    "import neo  # neo for some direct reading to check on # of segments etc\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NeuralynxIO: data/BiconditionalOdor/M040-2020-04-29-CDOD12\n",
      "nb_block: 1\n",
      "nb_segment:  [5]\n",
      "signal_channels: [CSC10, CSC11, CSC12, CSC13 ... CSC30 CSC31 CSC32 CSC9]\n",
      "unit_channels: [chTT3#13#0, chTT3#14#0, chTT3#8#0, chTT3#24#0 ... chTT8#3#0 chTT8#4#0 chTT8#19#0 chTT8#20#0]\n",
      "event_channels: [Events event_id=11 ttl=0, Events event_id=11 ttl=1, Events event_id=11 ttl=2, Events event_id=11 ttl=4 ... Events event_id=11 ttl=48 Events event_id=11 ttl=64 Events event_id=11 ttl=96 Events event_id=19 ttl=0]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# sample_file can be any file in the directory which has ALL the ncs files.\n",
    "# sample_file = 'data/BiconditionalOdor/M040-2020-04-28-CDOD11/CSC11.ncs'\n",
    "sample_file = 'data/BiconditionalOdor/M040-2020-04-29-CDOD12/CSC11.ncs'\n",
    "dname = (op.dirname(sample_file))\n",
    "# print(dname)\n",
    "\n",
    "# Throws an error if the the ncs files don't have the same length\n",
    "# On omitting the file with a different length, this works (data/BiconditionalOdor/M040-2020-04-28-CDOD11/CSC26.ncs for instance)\n",
    "# Or use the \"exclude_files\" keyword to explicitly a list of .Ncs files that are not meant for spike sorting or other_reasons\n",
    "# NOTE: Clone and install neo from \"https://github.com/manimoh/python-neo\" for the 'exclude_files' feature to work\n",
    "\n",
    "# reader = neo.NeuralynxIO(dirname=dname, exclude_files = ['CSC26.ncs'])\n",
    "reader = neo.NeuralynxIO(dirname=dname)\n",
    "print(reader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# recordingList = [se.NeuralynxRecordingExtractor(dirname=op.dirname(sample_file), exclude_files = ['CSC26.ncs'], seg_index=i) for i in range(reader.segment_count(0))]\n",
    "recordingList = [se.NeuralynxRecordingExtractor(dirname=op.dirname(sample_file), seg_index=i) for i in range(reader.segment_count(0))]\n",
    "\n",
    "# This concatenates mutiple segments together\n",
    "recordingFull = se.MultiRecordingTimeExtractor(recordingList)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<table><tr>\n",
    "<td> <img src=\"ks2_test/A4x2-tet-CM42-channelMap.png\" alt=\"Drawing\"/> </td>\n",
    "<td> <img src=\"ks2_test/A4x2-tet-CM42-probeGeometry.png\" alt=\"Drawing\"/> </td>\n",
    "</tr></table>\n",
    "<table><tr>\n",
    "<td> <img src=\"ks2_test/CM32-NN-to-neuralynx-mapping.JPG\" alt=\"Drawing\"/> </td>\n",
    "</tr></table>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load probe file\n",
    "recording_probe = recordingFull.load_probe_file('probe_files/NN_A4x2tet.prb')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Adding labels (CSC names) to channels from neo Neuralynx header\n",
    "for ad_ch in recording_probe.get_channel_ids():\n",
    "    csc_idx = np.where(reader.header['signal_channels']['id'] == ad_ch)\n",
    "#     print(reader.header['signal_channels']['name'][csc_idx])\n",
    "    recording_probe.set_channel_property(ad_ch, 'label', reader.header['signal_channels']['name'][csc_idx][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Assigning groups (TT num) to channels from neo Neuralynx header\n",
    "for ch in reader.header['unit_channels']:\n",
    "    tetrode = int(re.search('(?<=TT)(.*?)(?=#)', ch[0]).group(0))\n",
    "    ad_ch = int(re.search('(?<=#)(.*?)(?=#)', ch[0]).group(0))\n",
    "    recording_probe.set_channel_property(ad_ch, 'group', tetrode)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Sanity check to see that the relative locations of the various channels match that of the images above\n",
    "xcoords = [pos[0] for pos in recording_probe.get_channel_locations()]\n",
    "ycoords = [pos[1] for pos in recording_probe.get_channel_locations()]\n",
    "labels = [recording_probe.get_channel_property(i, 'label') for i in recording_probe.get_channel_ids()]\n",
    "fig,ax = plt.subplots()\n",
    "ax.axis('off')\n",
    "for i in range(len(xcoords)):\n",
    "    plt.text(xcoords[i]/200,ycoords[i]/200,labels[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Transfer properties from recording probe to recordingFull\n",
    "for ch in recordingFull.get_channel_ids():\n",
    "    for prop in recording_probe.get_shared_channel_property_names():\n",
    "        recordingFull.set_channel_property(ch, prop, recording_probe.get_channel_property(ch, prop))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['hdsort', 'herdingspikes', 'ironclust', 'kilosort', 'kilosort2', 'klusta', 'mountainsort4', 'spykingcircus', 'tridesclous', 'waveclus']\n"
     ]
    }
   ],
   "source": [
    "# This should contain kilosort2\n",
    "print (ss.available_sorters())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'detect_threshold': 5, 'projection_threshold': [10, 4], 'preclust_threshold': 8, 'car': True, 'minFR': 0.1, 'minfr_goodchannels': 0.1, 'freq_min': 150, 'sigmaMask': 30, 'nPCs': 3, 'ntbuff': 64, 'nfilt_factor': 4, 'NT': None, 'keep_good_only': False}\n"
     ]
    }
   ],
   "source": [
    "# These params should be changed and passed on kwargs if needed\n",
    "ks2_params = ss.Kilosort2Sorter.default_params()\n",
    "print(ks2_params)\n",
    "\n",
    "# Example of changing parameter\n",
    "ks2_params['projection_threshold'] = [20, 15]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nTo use Kilosort2 run:\\n\\n        >>> git clone https://github.com/MouseLand/Kilosort2\\n    and provide the installation path by setting the KILOSORT2_PATH\\n    environment variables or using Kilosort2Sorter.set_kilosort2_path().\\n\\n\\n\\n    More information on Kilosort2 at:\\n        https://github.com/MouseLand/Kilosort2\\n    '"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This displays how to set KS2 path\n",
    "ss.Kilosort2Sorter.installation_mesg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting KILOSORT2_PATH environment variable for subprocess calls to: D:\\AutomaticSpikeSort\\Kilosort2\n"
     ]
    }
   ],
   "source": [
    "# Set the path tothe Kilosort repository\n",
    "ss.Kilosort2Sorter.set_kilosort2_path('D:\\AutomaticSpikeSort\\Kilosort2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This should return true if the path is set correctly\n",
    "# Kilosort2 won't run if this is False\n",
    "ss.Kilosort2Sorter.is_installed()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING! The recording contains several group. In order to spike sort by 'group' use grouping_property='group' as argument.\n",
      "RUNNING SHELL SCRIPT: C:\\Users\\mvdmlab\\ss_ks2_test\\run_kilosort2.bat\n",
      "\n",
      "\n",
      "C:\\Users\\mvdmlab>cd C:\\Users\\mvdmlab\\ss_ks2_test \n",
      "\n",
      "\n",
      "\n",
      "C:\\Users\\mvdmlab\\ss_ks2_test>matlab -nosplash -wait -log -r kilosort2_master \n",
      "\n",
      "Time   0s. Determining good channels.. \n",
      "\n",
      "found 9934 threshold crossings in 214.29 seconds of data \n",
      "\n",
      "found 0 bad channels \n",
      "\n",
      "Time   9s. Computing whitening matrix.. \n",
      "\n",
      "Getting channel whitening matrix... \n",
      "\n",
      "Channel-whitening matrix computed. \n",
      "\n",
      "Time  14s. Loading raw data and applying filters... \n",
      "\n",
      "Time 146s. Finished preprocessing 2815 batches. \n",
      "\n",
      "random seed for clusterSingleBatches: 1\n",
      "\n",
      "Obtained 7 PC waveforms in 1.40 seconds \n",
      "\n",
      "time 0.25, pre clustered 1 / 2815 batches \n",
      "\n",
      "time 21.98, pre clustered 501 / 2815 batches \n",
      "\n",
      "time 43.31, pre clustered 1001 / 2815 batches \n",
      "\n",
      "time 64.96, pre clustered 1501 / 2815 batches \n",
      "\n",
      "time 86.72, pre clustered 2001 / 2815 batches \n",
      "\n",
      "time 108.25, pre clustered 2501 / 2815 batches \n",
      "\n",
      "time 0.22, compared 1 / 2815 batches \n",
      "\n",
      "time 3.40, compared 501 / 2815 batches \n",
      "\n",
      "time 6.57, compared 1001 / 2815 batches \n",
      "\n",
      "time 9.75, compared 1501 / 2815 batches \n",
      "\n",
      "time 12.94, compared 2001 / 2815 batches \n",
      "\n",
      "time 16.12, compared 2501 / 2815 batches \n",
      "\n",
      "time 25.44, Re-ordered 2815 batches. \n",
      "\n",
      "Time  30s. Optimizing templates ...\n",
      "\n",
      "30.49 sec, 1 / 2816 batches, 5 units, nspks: 1.2215, mu: 10.0000, nst0: 3, merges: 0.0000, 0.0000 \n",
      "\n",
      "34.65 sec, 101 / 2816 batches, 18 units, nspks: 17.2473, mu: 31.0997, nst0: 4, merges: 9.6224, 0.0000 \n",
      "\n",
      "38.35 sec, 201 / 2816 batches, 12 units, nspks: 17.4681, mu: 31.0843, nst0: 13, merges: 11.6243, 0.0900 \n",
      "\n",
      "41.91 sec, 301 / 2816 batches, 9 units, nspks: 19.4831, mu: 29.2214, nst0: 33, merges: 11.3935, 0.0109 \n",
      "\n",
      "45.49 sec, 401 / 2816 batches, 10 units, nspks: 20.1605, mu: 25.5156, nst0: 23, merges: 10.8685, 0.0013 \n",
      "\n",
      "49.03 sec, 501 / 2816 batches, 9 units, nspks: 17.5015, mu: 23.8469, nst0: 14, merges: 10.7235, 0.0002 \n",
      "\n",
      "52.57 sec, 601 / 2816 batches, 11 units, nspks: 19.8494, mu: 24.6762, nst0: 19, merges: 11.3382, 0.0000 \n",
      "\n",
      "56.13 sec, 701 / 2816 batches, 10 units, nspks: 20.2892, mu: 27.7745, nst0: 11, merges: 11.8012, 0.0000 \n",
      "\n",
      "59.66 sec, 801 / 2816 batches, 11 units, nspks: 20.0780, mu: 24.3805, nst0: 11, merges: 10.8010, 0.0000 \n",
      "\n",
      "63.21 sec, 901 / 2816 batches, 11 units, nspks: 26.0930, mu: 34.5195, nst0: 63, merges: 11.1264, 0.0000 \n",
      "\n",
      "66.86 sec, 1001 / 2816 batches, 11 units, nspks: 22.4436, mu: 24.3742, nst0: 15, merges: 12.0246, 0.0000 \n",
      "\n",
      "70.45 sec, 1101 / 2816 batches, 14 units, nspks: 24.5078, mu: 30.3028, nst0: 32, merges: 11.9607, 0.0000 \n",
      "\n",
      "74.26 sec, 1201 / 2816 batches, 15 units, nspks: 40.6529, mu: 31.9466, nst0: 86, merges: 12.1434, 0.0000 \n",
      "\n",
      "78.17 sec, 1301 / 2816 batches, 20 units, nspks: 78.2305, mu: 35.0437, nst0: 46, merges: 12.3609, 0.0000 \n",
      "\n",
      "82.89 sec, 1401 / 2816 batches, 14 units, nspks: 424.8376, mu: 32.3027, nst0: 224, merges: 14.1988, 0.0000 \n",
      "\n",
      "87.64 sec, 1501 / 2816 batches, 26 units, nspks: 169.3192, mu: 33.2209, nst0: 149, merges: 13.8963, 0.0000 \n",
      "\n",
      "91.85 sec, 1601 / 2816 batches, 21 units, nspks: 49.9893, mu: 34.2093, nst0: 104, merges: 13.0331, 0.0000 \n",
      "\n",
      "95.70 sec, 1701 / 2816 batches, 12 units, nspks: 28.4270, mu: 32.0262, nst0: 13, merges: 12.1780, 0.0000 \n",
      "\n",
      "99.39 sec, 1801 / 2816 batches, 12 units, nspks: 26.7721, mu: 24.9974, nst0: 14, merges: 12.0886, 0.0000 \n",
      "\n",
      "103.01 sec, 1901 / 2816 batches, 12 units, nspks: 26.2060, mu: 24.6622, nst0: 57, merges: 11.6195, 0.0000 \n",
      "\n",
      "106.64 sec, 2001 / 2816 batches, 18 units, nspks: 25.5013, mu: 28.3959, nst0: 54, merges: 10.0295, 0.0000 \n",
      "\n",
      "110.25 sec, 2101 / 2816 batches, 9 units, nspks: 20.8616, mu: 24.8368, nst0: 16, merges: 10.6324, 0.0000 \n",
      "\n",
      "113.86 sec, 2201 / 2816 batches, 14 units, nspks: 21.8947, mu: 28.9771, nst0: 31, merges: 11.1011, 0.0000 \n",
      "\n",
      "117.42 sec, 2301 / 2816 batches, 11 units, nspks: 23.2874, mu: 24.7200, nst0: 30, merges: 10.7197, 0.0000 \n",
      "\n",
      "120.90 sec, 2401 / 2816 batches, 11 units, nspks: 21.2800, mu: 24.6109, nst0: 47, merges: 10.0112, 0.0000 \n",
      "\n",
      "124.43 sec, 2501 / 2816 batches, 12 units, nspks: 21.4088, mu: 27.5667, nst0: 18, merges: 11.1712, 0.0000 \n",
      "\n",
      "128.04 sec, 2601 / 2816 batches, 10 units, nspks: 20.5143, mu: 28.3146, nst0: 21, merges: 11.3166, 0.0000 \n",
      "\n",
      "131.67 sec, 2701 / 2816 batches, 10 units, nspks: 18.5063, mu: 24.4623, nst0: 16, merges: 11.7537, 0.0000 \n",
      "\n",
      "135.23 sec, 2801 / 2816 batches, 12 units, nspks: 18.5742, mu: 24.9532, nst0: 30, merges: 11.1366, 0.0000 \n",
      "\n",
      "Elapsed time is 135.817307 seconds.\n",
      "\n",
      "Finished learning templates \n",
      "\n",
      "Time 136s. Optimizing templates ...\n",
      "\n",
      "136.18 sec, 1 / 1408 batches, 8 units, nspks: 0.7500, mu: 27.2469, nst0: 21 \n",
      "\n",
      "137.94 sec, 101 / 1408 batches, 8 units, nspks: 20.0324, mu: 27.9249, nst0: 13 \n",
      "\n",
      "139.69 sec, 201 / 1408 batches, 8 units, nspks: 18.7991, mu: 28.2673, nst0: 20 \n",
      "\n",
      "141.45 sec, 301 / 1408 batches, 8 units, nspks: 21.4732, mu: 28.1036, nst0: 56 \n",
      "\n",
      "143.22 sec, 401 / 1408 batches, 8 units, nspks: 20.5677, mu: 28.4427, nst0: 28 \n",
      "\n",
      "144.98 sec, 501 / 1408 batches, 8 units, nspks: 17.6445, mu: 28.2640, nst0: 16 \n",
      "\n",
      "146.74 sec, 601 / 1408 batches, 8 units, nspks: 19.7909, mu: 28.5416, nst0: 27 \n",
      "\n",
      "148.52 sec, 701 / 1408 batches, 8 units, nspks: 19.6001, mu: 28.5400, nst0: 15 \n",
      "\n",
      "150.32 sec, 801 / 1408 batches, 8 units, nspks: 19.7182, mu: 28.8940, nst0: 17 \n",
      "\n",
      "152.10 sec, 901 / 1408 batches, 8 units, nspks: 25.2261, mu: 29.9798, nst0: 71 \n",
      "\n",
      "153.86 sec, 1001 / 1408 batches, 8 units, nspks: 22.2858, mu: 30.7126, nst0: 20 \n",
      "\n",
      "155.61 sec, 1101 / 1408 batches, 8 units, nspks: 23.4112, mu: 31.5115, nst0: 33 \n",
      "\n",
      "157.37 sec, 1201 / 1408 batches, 8 units, nspks: 26.3020, mu: 34.8829, nst0: 15 \n",
      "\n",
      "159.13 sec, 1301 / 1408 batches, 8 units, nspks: 25.6839, mu: 36.1954, nst0: 41 \n",
      "\n",
      "160.87 sec, 1401 / 1408 batches, 8 units, nspks: 15.9615, mu: 37.9017, nst0: 8 \n",
      "\n",
      "Elapsed time is 160.989727 seconds.\n",
      "\n",
      "Time 161s. Optimizing templates ...\n",
      "\n",
      "161.08 sec, 1 / 1407 batches, 8 units, nspks: 0.9000, mu: 27.2469, nst0: 21 \n",
      "\n",
      "162.80 sec, 101 / 1407 batches, 8 units, nspks: 19.9944, mu: 27.8379, nst0: 23 \n",
      "\n",
      "164.54 sec, 201 / 1407 batches, 8 units, nspks: 19.3231, mu: 28.9689, nst0: 32 \n",
      "\n",
      "166.25 sec, 301 / 1407 batches, 8 units, nspks: 15.9478, mu: 29.2566, nst0: 13 \n",
      "\n",
      "167.96 sec, 401 / 1407 batches, 8 units, nspks: 16.6968, mu: 29.1000, nst0: 31 \n",
      "\n",
      "169.68 sec, 501 / 1407 batches, 8 units, nspks: 16.6166, mu: 28.8509, nst0: 14 \n",
      "\n",
      "171.43 sec, 601 / 1407 batches, 8 units, nspks: 16.8388, mu: 28.4816, nst0: 25 \n",
      "\n",
      "173.17 sec, 701 / 1407 batches, 8 units, nspks: 16.1691, mu: 29.1292, nst0: 25 \n",
      "\n",
      "174.92 sec, 801 / 1407 batches, 8 units, nspks: 16.6229, mu: 28.7584, nst0: 21 \n",
      "\n",
      "176.66 sec, 901 / 1407 batches, 8 units, nspks: 16.0020, mu: 29.1000, nst0: 36 \n",
      "\n",
      "178.38 sec, 1001 / 1407 batches, 8 units, nspks: 16.9487, mu: 29.5895, nst0: 22 \n",
      "\n",
      "180.11 sec, 1101 / 1407 batches, 8 units, nspks: 15.4345, mu: 29.7067, nst0: 14 \n",
      "\n",
      "181.84 sec, 1201 / 1407 batches, 8 units, nspks: 16.3135, mu: 29.9061, nst0: 21 \n",
      "\n",
      "183.55 sec, 1301 / 1407 batches, 8 units, nspks: 18.2212, mu: 33.1951, nst0: 14 \n",
      "\n",
      "185.29 sec, 1401 / 1407 batches, 8 units, nspks: 23.3215, mu: 40.6451, nst0: 48 \n",
      "\n",
      "Elapsed time is 185.396875 seconds.\n",
      "\n",
      "Number of spikes before applying cutoff: 69012\n",
      "\n",
      "Finished compressing time-varying templates \n",
      "\n",
      "initialized spike counts\n",
      "\n",
      "Found 1 splits, checked 1/9 clusters, nccg 0 \n",
      "\n",
      "Finished splitting. Found 5 splits, checked 13/13 clusters, nccg 0 \n",
      "\n",
      "Found 0 splits, checked 1/13 clusters, nccg 0 \n",
      "\n",
      "Finished splitting. Found 0 splits, checked 13/13 clusters, nccg 1 \n",
      "\n",
      "found 2 good units \n",
      "\n",
      "Saving results to Phy  \n",
      "\n",
      "kilosort2 run time 474.60s\n"
     ]
    }
   ],
   "source": [
    "# Does not work if the working/output folder is in a different drive/directory, must change the directory outside the function call\n",
    "# SS folks suggest the same, see https://github.com/SpikeInterface/spikesorters/issues/94\n",
    "# KS2 folks suggest that the output/Working directory should reside on an SSD\n",
    "\n",
    "# For now run KS2 on all channels without caring about channel groups. For discussion about details see: https://github.com/SpikeInterface/spikesorters/issues/150\n",
    "cwd = os.getcwd()\n",
    "os.chdir('C:/Users/mvdmlab')# ks2_out0 = ss.run_kilosort2(recordingFull)\n",
    "\n",
    "# This launches a matlab instance using a shell command so the KS2 log is present only on the matlab console, which vanishes after finishing its run or if it runs into an error.\n",
    "# Ideally the shell command needs to be modified so the Maltab console output is logged to somewhere more permanent.\n",
    "# Possible solutions below (Need to explore)\n",
    "# 1) https://www.mathworks.com/matlabcentral/answers/102082-how-do-i-call-matlab-from-the-dos-prompt#answer_419655\n",
    "# 2) https://stackoverflow.com/questions/7714709/writing-log-statements-to-standard-output-with-matlab/41818741#41818741\n",
    "\n",
    "# UPDATE: Opened an issue suggesting the same to SS folks here, https://github.com/SpikeInterface/spikesorters/issues/151\n",
    "\n",
    "ks2_out0 = ss.run_kilosort2(recordingFull, output_folder = 'C:/Users/mvdmlab/ss_ks2_test', **ks2_params, verbose=True)\n",
    "\n",
    "# TODO: Delete the recording.dat file and the temp.wh file since they are needed only for the sorting and not later on for comapring/viewing results and occupy a lot of space\n",
    "\n",
    "# Alternate way of runnng the kilosort. Works!!\n",
    "# ks2_out1 = ss.run_sorters(['kilosort2'],[recordingFull], working_folder='C:/Users/mvdmlab/ss_ks2test')\n",
    "\n",
    "os.chdir(cwd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This runs KS2 separately on each channel group. It creates one folder for each channel group, with the folder names starting\n",
    "# from 0 to the number of groups-1.\n",
    "# This is not recommended by the KS2 folks: https://github.com/MouseLand/Kilosort2/issues/21\n",
    "# Works!!\n",
    "# ks2_out2 = ss.run_kilosort2(recordingFull, output_folder = 'C:/Users/mvdmlab/ss_ks2test_group', grouping_property = 'group')"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}