src/Nipype/Notebooks/spm_event_categorical.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Categorical Modelling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nilearn import plotting\n",
    "%matplotlib inline\n",
    "import os\n",
    "import json\n",
    "from nipype.interfaces import fsl \n",
    "from nipype.interfaces import spm\n",
    "from nipype.interfaces.spm import (Realign, SliceTiming, Coregister,  NewSegment,  Normalize12, Smooth)\n",
    "from nipype.interfaces.spm import Level1Design, EstimateModel, EstimateContrast\n",
    "from nipype.algorithms.modelgen import SpecifySPMModel\n",
    "from nipype.interfaces import matlab as mlab\n",
    "from nipype.interfaces.io import SelectFiles, DataSink\n",
    "import nipype.interfaces.utility as util \n",
    "from nipype.algorithms import rapidart as ra\n",
    "from nipype.interfaces.utility import Function, IdentityInterface\n",
    "import nipype.pipeline.engine as pe\n",
    "import nipype.interfaces.io as nio\n",
    "from nipype.interfaces.base import Bunch\n",
    "from nipype import DataGrabber, Workflow, Node\n",
    "from scipy.io.matlab import loadmat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# necessary to let nipype know about matlab path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "stty: 'standard input': Inappropriate ioctl for device\n"
     ]
    }
   ],
   "source": [
    "spm.SPMCommand.set_mlab_paths(paths=os.path.abspath(os.path.join(os.environ['HOME'], 'Documents/MATLAB/spm12/')), matlab_cmd='/soft/matlab_hd/R2020b/bin/glnxa64/MATLAB -nodesktop -nosplash')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "mlab.MatlabCommand.set_default_matlab_cmd(\"/soft/matlab_hd/R2020b/bin/glnxa64/MATLAB  -nodesktop -nosplash\")\n",
    "mlab.MatlabCommand.set_default_paths(os.path.abspath(os.path.join(os.environ['HOME'], 'Documents/MATLAB/spm12/')))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# spm.SPMCommand().version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "fsl.FSLCommand.set_default_output_type('NIFTI')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_dir = os.path.join(os.environ['HOME'], 'spmbasics/data/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "experiment_dir = os.path.join(base_dir, 'output')\n",
    "data_dir = os.path.abspath(os.path.join(base_dir, 'face_rep'))\n",
    "output_dir = 'nipype'\n",
    "working_dir = 'workingdir'\n",
    "\n",
    "# list of subject identifiers\n",
    "subject_list = ['M03953']\n",
    "\n",
    "# TR of functional images\n",
    "TR = 2.\n",
    "\n",
    "\n",
    "# Smoothing width used during preprocessing\n",
    "fwhm = [8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# add how to refer sots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#mat0 = mat = loadmat(os.path.join(data_dir, \"sots.mat\"), mat_dtype=True, matlab_compatible=True, struct_as_record=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "NOT WORKING: https://nipype.readthedocs.io/en/latest/users/examples/fmri_spm_face.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from scipy.io.matlab import loadmat\n",
    "#mat = loadmat(os.path.join(data_dir, \"sots.mat\"), struct_as_record=False)\n",
    "#sot = mat['sot'][0]\n",
    "#itemlag = mat['itemlag'][0]\n",
    "\n",
    "#subjectinfo = [\n",
    "#    Bunch(\n",
    "#        conditions=['N1', 'N2', 'F1', 'F2'],\n",
    "#        onsets=[sot[0], sot[1], sot[2], sot[3]],\n",
    "#        durations=[[0], [0], [0], [0]],\n",
    "#        amplitudes=None,\n",
    "#        tmod=None,\n",
    "#        pmod=None,\n",
    "#        regressor_names=None,\n",
    "#        regressors=None)\n",
    "#]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "mat = loadmat(os.path.join(data_dir, \"sots.mat\"), mat_dtype=True, matlab_compatible=True, struct_as_record=True, simplify_cells=True)\n",
    "#sot = mat['sot'][1]\n",
    "\n",
    "#itemlag = mat['itemlag'][0]\n",
    "#onsets=[sot[0], sot[1], sot[2], sot[3]],\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "mat = loadmat(os.path.join(data_dir, \"sots.mat\"), mat_dtype=True, matlab_compatible=True, struct_as_record=True, simplify_cells=True)\n",
    "subjectinfo = [\n",
    "    Bunch(\n",
    "        conditions=['N1', 'N2', 'F1', 'F2'],\n",
    "        onsets=[ mat['sot'][0], mat['sot'][1],  mat['sot'][2],  mat['sot'][3]],\n",
    "        durations=[[0], [0], [0], [0]],\n",
    "        amplitudes=None,\n",
    "        tmod=None,\n",
    "        pmod=None,\n",
    "        regressor_names=None,\n",
    "        regressors=None)\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# design matrix setting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "cond1 = ('positive effect of condition', 'T',\n",
    "         ['N1', 'N2', 'F1', 'F2'], [1, 1, 1, 1])\n",
    "cond2 = ('positive effect of condition_dtemo', 'T',\n",
    "          ['N1', 'N2', 'F1', 'F2'], [1, 1, 1, 1])\n",
    "cond3 = ('positive effect of condition_ddisp', 'T',\n",
    "          ['N1', 'N2', 'F1', 'F2'], [1, 1, 1, 1])\n",
    "# non-famous > famous\n",
    "fam1 = ('positive effect of Fame', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, 1, -1, -1])\n",
    "fam2 = ('positive effect of Fame_dtemp', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, 1, -1, -1])\n",
    "fam3 = ('positive effect of Fame_ddisp', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, 1, -1, -1])\n",
    "# rep1 > rep2\n",
    "rep1 = ('positive effect of Rep', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, -1, 1, -1])\n",
    "rep2 = ('positive effect of Rep_dtemp', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, -1, 1, -1])\n",
    "rep3 = ('positive effect of Rep_ddisp', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, -1, 1, -1])\n",
    "int1 = ('positive interaction of Fame x Rep', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [-1, -1, -1, 1])\n",
    "int2 = ('positive interaction of Fame x Rep_dtemp', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, -1, -1, 1])\n",
    "int3 = ('positive interaction of Fame x Rep_ddisp', 'T',\n",
    "        ['N1', 'N2', 'F1', 'F2'], [1, -1, -1, 1])\n",
    "\n",
    "contf1 = ['average effect condition', 'F', [cond1, cond2, cond3]]\n",
    "contf2 = ['main effect Fam', 'F', [fam1, fam2, fam3]]\n",
    "contf3 = ['main effect Rep', 'F', [rep1, rep2, rep3]]\n",
    "contf4 = ['interaction: Fam x Rep', 'F', [int1, int2, int3]]\n",
    "contrast_list = [\n",
    "    cond1, cond2, cond3, fam1, fam2, fam3, rep1, rep2, rep3, int1, int2, int3,\n",
    "    contf1, contf2, contf3, contf4\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SpecifyModel - Generates SPM-specific Model\n",
    "specifymodel = Node(SpecifySPMModel(concatenate_runs=False,\n",
    "                                 input_units='scans',\n",
    "                                 output_units='scans',\n",
    "                                 time_repetition=TR,\n",
    "                                 high_pass_filter_cutoff=128,\n",
    "                                 subject_info = subjectinfo),\n",
    "                 name=\"specifymodel\")\n",
    "\n",
    "# Level1Design - Generates an SPM design matrix same as the first level tutorial\n",
    "eventcategorical = Node(Level1Design(bases={'hrf': {'derivs': [1, 1]}},\n",
    "                                 timing_units='scans',\n",
    "                                 interscan_interval=TR,\n",
    "                                 volterra_expansion_order=1, # no model interction\n",
    "                                 flags={'mthresh': 0.8},\n",
    "                                 global_intensity_normalization='none',\n",
    "                                 microtime_onset=12,\n",
    "                                 microtime_resolution=24,\n",
    "                                 #factor_info = [dict(name = 'Fame', levels = 2), # figure out factorized output names\n",
    "                                 #               dict(name = 'Rep', levels = 2)],\n",
    "                                 model_serial_correlations='AR(1)'), #matlabbatch{1}.spm.stats.fmri_spec.cvi = 'AR(1)';\n",
    "                    name=\"eventcategorical\")\n",
    "\n",
    "# EstimateModel - estimate the parameters of the model\n",
    "categoricalestimate = Node(EstimateModel(estimation_method={'Classical': 1}),\n",
    "                      write_residuals=False, \n",
    "                      name=\"categoricalestimate\")\n",
    "\n",
    "# EstimateContrast - estimates contrasts\n",
    "categoricalconest = Node(EstimateContrast(contrasts = contrast_list),\n",
    "                    use_derivs=True, \n",
    "                    name=\"categoricalconest\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Infosource - a function free node to iterate over the list of subject names\n",
    "infosource = Node(IdentityInterface(fields=['subject_id',\n",
    "                                            'contrasts'],\n",
    "                                    contrasts=contrast_list),\n",
    "                  name=\"infosource\")\n",
    "infosource.iterables = [('subject_id', subject_list)]\n",
    "\n",
    "\n",
    "templates = {'func': os.path.join(output_dir, 'event_preproc', '_subject_id_{subject_id}',\n",
    "                         'swars{subject_id}_0005_*.img'),#sM03953_0005_*.img\n",
    "             'mc_param': os.path.join(output_dir, 'event_preproc', '_subject_id_{subject_id}',\n",
    "                         'rp_s{subject_id}_0005_0006.txt')}\n",
    "selectfiles = Node(SelectFiles(templates,\n",
    "                               base_directory=experiment_dir,\n",
    "                               sort_filelist=True),\n",
    "                   name=\"selectfiles\")\n",
    "\n",
    "# Datasink - creates output folder for important outputs\n",
    "datasink = Node(DataSink(base_directory=experiment_dir,\n",
    "                         container=output_dir),\n",
    "                name=\"datasink\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def makelist(item):\n",
    "    return [item] # look how similar thing incoorporated."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initiation of the 1st-level analysis workflow\n",
    "event_cat = Workflow(name='event_cat')\n",
    "event_cat.base_dir = os.path.join(experiment_dir, working_dir)\n",
    "\n",
    "# Connect up the 1st-level analysis components\n",
    "event_cat.connect([(infosource, selectfiles, [('subject_id', 'subject_id')]),\n",
    "                    (infosource, categoricalconest, [('contrasts', 'contrasts')]),\n",
    "                    (selectfiles, specifymodel, [(('func', makelist),  'functional_runs')]), # [(('smooth.smoothed_files',makelist), 'modelspec.functional_runs')])])\n",
    "\n",
    "                    (selectfiles, specifymodel, [('mc_param', 'realignment_parameters')]),\n",
    "                    (specifymodel, eventcategorical, [('session_info','session_info')]),\n",
    "                    (eventcategorical, categoricalestimate, [('spm_mat_file','spm_mat_file')]),                    \n",
    "                    (categoricalestimate, categoricalconest, [('spm_mat_file','spm_mat_file'),\n",
    "                                                              ('beta_images','beta_images'),\n",
    "                                                              ('residual_image', 'residual_image')]),\n",
    "                    (categoricalestimate, datasink, [('mask_image','event_cat.@mask_image'),\n",
    "                                                    ('beta_images','event_cat.@beta_images'),\n",
    "                                                    ('residual_images','event_cat.@residual_images'),\n",
    "                                                    ('residual_image','event_cat.@residual_image'),\n",
    "                                                    ('RPVimage', 'event_cat.@RPVimage')]),\n",
    "                    (categoricalconest, datasink, [('spm_mat_file', 'event_cat.@spm_mat'),\n",
    "                                                   ('spmT_images', 'event_cat.@T'),\n",
    "                                                   ('con_images', 'event_cat.@con'),\n",
    "                                                   ('spmF_images', 'event_cat.@F'),\n",
    "                                                   ('ess_images', 'event_cat.@ess')]),\n",
    "                    ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:41:56,222 nipype.workflow INFO:\n",
      "\t Generated workflow graph: /home/matay/spmbasics/data/output/workingdir/event_cat/colored_cat.png (graph2use=colored, simple_form=True).\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'/home/matay/spmbasics/data/output/workingdir/event_cat/colored_cat.png'"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "event_cat.write_graph(graph2use='colored', format='png', dotfilename='colored_cat.dot', simple_form=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:41:57,597 nipype.workflow INFO:\n",
      "\t Generated workflow graph: /home/matay/spmbasics/data/output/workingdir/event_cat/flat_cat.png (graph2use=flat, simple_form=True).\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'/home/matay/spmbasics/data/output/workingdir/event_cat/flat_cat.png'"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "event_cat.write_graph(graph2use='flat', format='png', dotfilename='flat_cat.dot', simple_form=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Visualize the graph\n",
    "from IPython.display import Image\n",
    "Image(filename=os.path.join(event_cat.base_dir, 'event_cat', 'colored_cat.png'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:42:02,390 nipype.workflow INFO:\n",
      "\t Workflow event_cat settings: ['check', 'execution', 'logging', 'monitoring']\n",
      "240528-10:42:02,397 nipype.workflow INFO:\n",
      "\t Running in parallel.\n",
      "240528-10:42:02,398 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 0 tasks, and 1 jobs ready. Free memory (GB): 13.77/13.77, Free processors: 4/4.\n",
      "240528-10:42:02,471 nipype.workflow INFO:\n",
      "\t [Node] Setting-up \"event_cat.selectfiles\" in \"/home/matay/spmbasics/data/output/workingdir/event_cat/_subject_id_M03953/selectfiles\".\n",
      "240528-10:42:02,479 nipype.workflow INFO:\n",
      "\t [Node] Executing \"selectfiles\" <nipype.interfaces.io.SelectFiles>\n",
      "240528-10:42:02,490 nipype.workflow INFO:\n",
      "\t [Node] Finished \"selectfiles\", elapsed time 0.008862s.\n",
      "240528-10:42:04,403 nipype.workflow INFO:\n",
      "\t [Job 0] Completed (event_cat.selectfiles).\n",
      "240528-10:42:04,408 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 0 tasks, and 1 jobs ready. Free memory (GB): 13.77/13.77, Free processors: 4/4.\n",
      "240528-10:42:04,480 nipype.workflow INFO:\n",
      "\t [Node] Setting-up \"event_cat.specifymodel\" in \"/home/matay/spmbasics/data/output/workingdir/event_cat/_subject_id_M03953/specifymodel\".\n",
      "240528-10:42:04,527 nipype.workflow INFO:\n",
      "\t [Node] Executing \"specifymodel\" <nipype.algorithms.modelgen.SpecifySPMModel>\n",
      "240528-10:42:04,530 nipype.workflow INFO:\n",
      "\t [Node] Finished \"specifymodel\", elapsed time 0.001367s.\n",
      "240528-10:42:06,404 nipype.workflow INFO:\n",
      "\t [Job 1] Completed (event_cat.specifymodel).\n",
      "240528-10:42:06,408 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 0 tasks, and 1 jobs ready. Free memory (GB): 13.77/13.77, Free processors: 4/4.\n",
      "240528-10:42:06,477 nipype.workflow INFO:\n",
      "\t [Node] Setting-up \"event_cat.eventcategorical\" in \"/home/matay/spmbasics/data/output/workingdir/event_cat/_subject_id_M03953/eventcategorical\".\n",
      "240528-10:42:06,511 nipype.workflow INFO:\n",
      "\t [Node] Executing \"eventcategorical\" <nipype.interfaces.spm.model.Level1Design>\n",
      "240528-10:42:08,405 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 1 tasks, and 0 jobs ready. Free memory (GB): 13.57/13.77, Free processors: 3/4.\n",
      "                     Currently running:\n",
      "                       * event_cat.eventcategorical\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "stty: 'standard input': Inappropriate ioctl for device\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:42:25,434 nipype.workflow INFO:\n",
      "\t [Node] Finished \"eventcategorical\", elapsed time 12.455252s.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "stty: 'standard input': Inappropriate ioctl for device\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:42:26,424 nipype.workflow INFO:\n",
      "\t [Job 2] Completed (event_cat.eventcategorical).\n",
      "240528-10:42:26,429 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 0 tasks, and 1 jobs ready. Free memory (GB): 13.77/13.77, Free processors: 4/4.\n",
      "240528-10:42:26,496 nipype.workflow INFO:\n",
      "\t [Node] Setting-up \"event_cat.categoricalestimate\" in \"/home/matay/spmbasics/data/output/workingdir/event_cat/_subject_id_M03953/categoricalestimate\".\n",
      "240528-10:42:26,502 nipype.workflow INFO:\n",
      "\t [Node] Executing \"categoricalestimate\" <nipype.interfaces.spm.model.EstimateModel>\n",
      "240528-10:42:28,425 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 1 tasks, and 0 jobs ready. Free memory (GB): 13.57/13.77, Free processors: 3/4.\n",
      "                     Currently running:\n",
      "                       * event_cat.categoricalestimate\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "stty: 'standard input': Inappropriate ioctl for device\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:42:49,577 nipype.workflow INFO:\n",
      "\t [Node] Finished \"categoricalestimate\", elapsed time 17.331593s.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "stty: 'standard input': Inappropriate ioctl for device\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:42:50,448 nipype.workflow INFO:\n",
      "\t [Job 3] Completed (event_cat.categoricalestimate).\n",
      "240528-10:42:50,452 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 0 tasks, and 1 jobs ready. Free memory (GB): 13.77/13.77, Free processors: 4/4.\n",
      "240528-10:42:50,510 nipype.workflow INFO:\n",
      "\t [Node] Setting-up \"event_cat.categoricalconest\" in \"/home/matay/spmbasics/data/output/workingdir/event_cat/_subject_id_M03953/categoricalconest\".\n",
      "240528-10:42:50,534 nipype.workflow INFO:\n",
      "\t [Node] Executing \"categoricalconest\" <nipype.interfaces.spm.model.EstimateContrast>\n",
      "240528-10:42:52,450 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 1 tasks, and 0 jobs ready. Free memory (GB): 13.57/13.77, Free processors: 3/4.\n",
      "                     Currently running:\n",
      "                       * event_cat.categoricalconest\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "stty: 'standard input': Inappropriate ioctl for device\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:43:10,412 nipype.workflow INFO:\n",
      "\t [Node] Finished \"categoricalconest\", elapsed time 12.873159s.\n",
      "240528-10:43:10,466 nipype.workflow INFO:\n",
      "\t [Job 4] Completed (event_cat.categoricalconest).\n",
      "240528-10:43:10,470 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 0 tasks, and 1 jobs ready. Free memory (GB): 13.77/13.77, Free processors: 4/4.\n",
      "240528-10:43:10,548 nipype.workflow INFO:\n",
      "\t [Node] Setting-up \"event_cat.datasink\" in \"/home/matay/spmbasics/data/output/workingdir/event_cat/_subject_id_M03953/datasink\".\n",
      "240528-10:43:10,558 nipype.workflow INFO:\n",
      "\t [Node] Executing \"datasink\" <nipype.interfaces.io.DataSink>\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "stty: 'standard input': Inappropriate ioctl for device\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240528-10:43:10,565 nipype.workflow INFO:\n",
      "\t [Node] Finished \"datasink\", elapsed time 0.005288s.\n",
      "240528-10:43:12,468 nipype.workflow INFO:\n",
      "\t [Job 5] Completed (event_cat.datasink).\n",
      "240528-10:43:12,472 nipype.workflow INFO:\n",
      "\t [MultiProc] Running 0 tasks, and 0 jobs ready. Free memory (GB): 13.77/13.77, Free processors: 4/4.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<networkx.classes.digraph.DiGraph at 0x7fb636006040>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "event_cat.run('MultiProc', plugin_args={'n_procs': 4})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}