def create_workflow_hrfpattern_spm():
# GLM
design = Node(interface=spm_design(), name='design_glm')
design.inputs.timing_units = 'secs'
design.inputs.interscan_interval = .85
design.inputs.bases = {'hrf': {'derivs': [0, 0]}}
estimate = Node(interface=EstimateModel(), name="estimate")
estimate.inputs.estimation_method = {'Classical': 1}
contrastestimate = Node(interface=EstimateContrast(), name="contrast")
contrastestimate.inputs.contrasts = [('Visual', 'T', [
'1',
], [
1,
])]
w = Workflow(name='hrfpattern_spm')
w.connect(input_node, 'bold', model, 'functional_runs')
w.connect(input_node, 'events', model, 'bids_event_file')
w.connect(model, 'session_info', design, 'session_info')
w.connect(design, 'spm_mat_file', estimate, 'spm_mat_file')
w.connect(estimate, 'spm_mat_file', contrastestimate, 'spm_mat_file')
w.connect(estimate, 'beta_images', contrastestimate, 'beta_images')
w.connect(estimate, 'residual_image', contrastestimate, 'residual_image')
w.connect(contrastestimate, 'spmT_images', output_node, 'T_image')
return w
def prepare_indT_cs():
outdirs = [
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_con_vs_stroke/alff_analysis/',
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_con_vs_stroke/falff_analysis/'
]
for n, outdir in enumerate(outdirs):
stroke_hdir = '/home/peter/Desktop/prepare/rest/output/'
all_stroke_files = glob.glob(stroke_hdir + '/*/f_alff/' + maps[n])
stroke_files = [x for x in all_stroke_files if '12months' not in x]
healthy_hdir = '/home/peter/Desktop/Connect/rest/output/'
all_healthy_files = glob.glob(healthy_hdir + '/D_H*/f_alff/' + maps[n])
healthy_files = [x for x in all_healthy_files if 'P2' not in x]
ttest = Node(TwoSampleTTestDesign(), name='TwoSampleT')
ttest.inputs.group1_files = healthy_files
ttest.inputs.group2_files = stroke_files
modelEst = Node(EstimateModel(), name='EstimateModel')
modelEst.inputs.estimation_method = {'Classical': 1}
conEst = Node(EstimateContrast(), name='EstimateContrasts')
con_1 = ('Controls', 'T', ['Group_{1}', 'Group_{2}'], [1.0, 0.0])
con_2 = ('Patients', 'T', ['Group_{1}', 'Group_{2}'], [0.0, 1.0])
con_3 = ('Controls>Patients', 'T', ['Group_{1}',
'Group_{2}'], [1.0, -1.0])
con_4 = ('Patients>Controls', 'T', ['Group_{1}',
'Group_{2}'], [-1.0, 1.0])
contrasts = [con_1, con_2, con_3, con_4]
conEst.inputs.contrasts = contrasts
conEst.inputs.group_contrast = True
l2analysis = Workflow(name='l2analysis')
l2analysis.base_dir = outdir
l2analysis.connect([
(ttest, modelEst, [('spm_mat_file', 'spm_mat_file')]),
(modelEst, conEst, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image', 'residual_image')]),
])
l2analysis.write_graph(graph2use='colored')
l2analysis.run('MultiProc', plugin_args={'n_procs': 1})
selectderivs = Node(SelectFiles(templates, sort_filelist=True),
name='selectderivs')
#selectderivs.inputs.sub_id = subs
# One Sample T-Test Design - creates one sample T-Test Design
onesamplettestdes = Node(
OneSampleTTestDesign(),
#overwrite=True,
name="onesampttestdes")
onesamplettestdes.inputs.explicit_mask_file = mask
# EstimateModel - estimate the parameters of the model
# Even for second level it should be 'Classical': 1.
level2estimate = Node(EstimateModel(estimation_method={'Classical': 1}),
name="level2estimate")
# EstimateContrast - estimates simple group contrast
level2conestimate = Node(EstimateContrast(group_contrast=True),
name="level2conestimate")
cont1 = ['Group', 'T', ['mean'], [1]]
level2conestimate.inputs.contrasts = [cont1]
## Create the 2nd level pipeline
secondlev = Workflow(name='secondlev', base_dir=out_dir + '/tmp')
secondlev.config['execution'][
'crashdump_dir'] = base_dir = out_dir + '/tmp/crash_files'
secondlev.connect([
(infosource, selectderivs, [('contrast_id', 'contrast_id')]),
(selectderivs, onesamplettestdes, [('cons', 'in_files')]),
(onesamplettestdes, level2estimate, [('spm_mat_file', 'spm_mat_file')]),
(level2estimate, level2conestimate, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
def batch_paramatric_GLM(nii_root_dir, sub_num_list, total_session_num,
all_sub_dataframe, params_name, contrast_list,
cache_folder, result_folder, parallel_cores):
from nipype import Node, Workflow, Function
from nipype.interfaces.spm import Level1Design, EstimateModel, EstimateContrast
from nipype.algorithms.modelgen import SpecifySPMModel
from nipype.interfaces.utility import IdentityInterface
from nipype import DataSink
# Define the helper functions
def nii_selector(root_dir,
sub_num,
session_num,
all_sub_dataframe,
data_type="Smooth_8mm"):
import os
import glob
session_list = ["session" + str(i) for i in range(1, session_num + 1)]
sub_name = "sub" + str(sub_num)
# print(file_path)
nii_list = []
for s in session_list:
file_path = os.path.join(root_dir, sub_name, data_type, s)
nii_list.append(glob.glob(file_path + "/*.nii"))
single_sub_data = all_sub_dataframe[all_sub_dataframe.Subject_num ==
sub_num]
return (nii_list, single_sub_data, sub_name)
def condition_generator(single_sub_data, params_name, duration=2):
from nipype.interfaces.base import Bunch
run_num = set(single_sub_data.run)
subject_info = []
for i in run_num:
tmp_table = single_sub_data[single_sub_data.run == i]
tmp_onset = tmp_table.onset.values.tolist()
pmod_names = []
pmod_params = []
pmod_poly = []
for param in params_name:
pmod_params.append(tmp_table[param].values.tolist())
pmod_names.append(param)
pmod_poly.append(1)
tmp_Bunch = Bunch(conditions=["trial_onset_run" + str(i)],
onsets=[tmp_onset],
durations=[[duration]],
pmod=[
Bunch(name=pmod_names,
poly=pmod_poly,
param=pmod_params)
])
subject_info.append(tmp_Bunch)
return subject_info
# Define each Nodes in the workflow
NiiSelector = Node(Function(
input_names=[
"root_dir", "sub_num", "session_num", "all_sub_dataframe",
"data_type"
],
output_names=["nii_list", "single_sub_data", "sub_name"],
function=nii_selector),
name="NiiSelector")
ConditionGenerator = Node(Function(
input_names=["single_sub_data", "params_name", "duration"],
output_names=["subject_info"],
function=condition_generator),
name="ConditionGenerator")
glm_input = Node(IdentityInterface(
fields=['nii_list', 'single_sub_data', 'params_name', 'contrast_list'],
mandatory_inputs=True),
name="glm_input")
# SpecifyModel - Generates SPM-specific Model
modelspec = Node(SpecifySPMModel(concatenate_runs=False,
input_units='scans',
output_units='scans',
time_repetition=2,
high_pass_filter_cutoff=128),
name="modelspec")
# Level1Design - Generates an SPM design matrix
level1design = Node(Level1Design(bases={'hrf': {
'derivs': [0, 0]
}},
timing_units='scans',
interscan_interval=2),
name="level1design")
level1estimate = Node(EstimateModel(estimation_method={'Classical': 1}),
name="level1estimate")
level1conest = Node(EstimateContrast(), name="level1conest")
OutputNode = Node(DataSink(), name="OutputNode")
# Define the attributes of those nodes
NiiSelector.inputs.root_dir = nii_root_dir
NiiSelector.iterables = ("sub_num", sub_num_list)
NiiSelector.inputs.session_num = total_session_num
NiiSelector.inputs.data_type = "Smooth_8mm"
NiiSelector.inputs.all_sub_dataframe = all_sub_dataframe
glm_input.inputs.params_name = params_name
glm_input.inputs.contrast_list = contrast_list
OutputNode.inputs.base_directory = result_folder
# Define the workflows
single_sub_GLM_wf = Workflow(name='single_sub_GLM_wf')
single_sub_GLM_wf.connect([
(glm_input, ConditionGenerator, [('single_sub_data',
'single_sub_data'),
('params_name', 'params_name')]),
(glm_input, modelspec, [('nii_list', 'functional_runs')]),
(glm_input, level1conest, [('contrast_list', 'contrasts')]),
(ConditionGenerator, modelspec, [('subject_info', 'subject_info')]),
(modelspec, level1design, [('session_info', 'session_info')]),
(level1design, level1estimate, [('spm_mat_file', 'spm_mat_file')]),
(level1estimate, level1conest, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image', 'residual_image')])
])
batch_GLM_wf = Workflow(name="batch_GLM_wf", base_dir=cache_folder)
batch_GLM_wf.connect([(NiiSelector, single_sub_GLM_wf, [
('nii_list', 'glm_input.nii_list'),
('single_sub_data', 'glm_input.single_sub_data')
]), (NiiSelector, OutputNode, [('sub_name', 'container')]),
(single_sub_GLM_wf, OutputNode,
[('level1conest.spm_mat_file', '1stLevel.@spm_mat'),
('level1conest.spmT_images', '1stLevel.@T'),
('level1conest.con_images', '1stLevel.@con'),
('level1conest.spmF_images', '1stLevel.@F'),
('level1conest.ess_images', '1stLevel.@ess')])])
# Excute the workflow
batch_GLM_wf.run(plugin='MultiProc',
plugin_args={'n_procs': parallel_cores})
interscan_interval=2.0,
microtime_resolution=16,
microtime_onset=1,
bases={'hrf': {
'derivs': [0, 0]
}},
global_intensity_normalization='none',
mask_threshold=0.8,
model_serial_correlations='AR(1)',
volterra_expansion_order=2),
name='model_spec')
est_model = Node(EstimateModel(estimation_method={'Classical': 1}),
name='est_model')
est_con = Node(EstimateContrast(), name='est_con')
## CREATE WORKFLOW
# Create a workflow to return the seed nuisance regressors and seed map(s) for a subject
first_level = Workflow(name='first_level')
first_level.base_dir = '/data/mridata/jdeng/tools/first_level/nipype'
# Datasink
substitutions = [('output', '')]
datasink = MapNode(DataSink(parameterization=False,
substitutions=substitutions),
name="datasink",
iterfield=['base_directory', 'container'])
# Helper functions for connections
name='level1design')
level1design.inputs.bases = {'hrf': {'derivs': [0, 0]}}
level1design.inputs.timing_units = 'secs'
level1design.inputs.interscan_interval = TR
firstlev.connect([
(modelspec, level1design, [('session_info', 'session_info')])
])
level1estimate = Node(EstimateModel(),
overwrite=False,
name='level1estimate')
level1estimate.inputs.estimation_method = {'Classical': 1}
firstlev.connect(level1design, 'spm_mat_file',
level1estimate, 'spm_mat_file')
contrast_estimate = Node(EstimateContrast(),
overwrite=False,
name='contraste_estimate')
contrast_estimate.config = {'execution': {'remove_unnecessary_outputs': False}}
firstlev.connect([
(level1estimate, contrast_estimate,
[('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image', 'residual_image')])
])
contrasts = Node(Function(function=con_setup,
input_names=['subject_id'],
output_names=['contrasts']),
name='contrasts')
firstlev.connect([
# Level1Design - Generates an SPM design matrix
level1design = Node(Level1Design(bases={'hrf': {
'derivs': [0, 0]
}},
timing_units='secs',
interscan_interval=TR,
model_serial_correlations='AR(1)'),
name="level1design")
# EstimateModel - estimate the parameters of the model
level1estimate = Node(EstimateModel(estimation_method={'Classical': 1}),
name="level1estimate")
# EstimateContrast - estimates contrasts
conestimate = Node(EstimateContrast(), name="conestimate")
# Volume Transformation - transform contrasts into anatomical space
applyVolReg = MapNode(ApplyVolTransform(fs_target=True),
name='applyVolReg',
iterfield=['source_file'])
# MRIConvert - to gzip output files
mriconvert = MapNode(MRIConvert(out_type='niigz'),
name='mriconvert',
iterfield=['in_file'])
# Initiation of the 1st-level analysis workflow
l1analysis = Workflow(name='l1analysis')
# Connect up the 1st-level analysis components
# Level1Design - Generates an SPM design matrix
level1design = Node(Level1Design(bases={'hrf': {
'derivs': [1, 0]
}},
timing_units='secs',
interscan_interval=TR,
model_serial_correlations='FAST'),
name="level1design")
# EstimateModel - estimate the parameters of the model
level1estimate = Node(EstimateModel(estimation_method={'Classical': 1}),
name="level1estimate")
# EstimateContrast - estimates contrasts
level1conest = Node(EstimateContrast(), name="level1conest")
# Condition names
condition_names = ['CueA', 'CueB']
# Contrasts
con_01 = ['average', 'T', condition_names, [1 / 2., 1 / 2.]]
con_02 = ['CueA', 'T', condition_names, [1, 0]]
con_03 = ['CueB', 'T', condition_names, [0, 1]]
con_04 = ['CueA > CueB', 'T', condition_names, [1, -1]]
con_05 = ['CueB > CueA', 'T', condition_names, [-1, 1]]
contrast_list = [con_01, con_02, con_03, con_04, con_05]
def subjectinfo(subject_id):
def first_level(TR,
contrast_list,
subject_list,
experiment_dir,
output_dir,
subjectinfo_func,
working_dir='workingdir'):
"""define first level model"""
# SpecifyModel - Generates SPM-specific Model
modelspec = Node(SpecifySPMModel(concatenate_runs=False,
input_units='secs',
output_units='secs',
time_repetition=TR,
high_pass_filter_cutoff=128),
name="modelspec")
# Level1Design - Generates an SPM design matrix
level1design = Node(Level1Design(bases={'hrf': {
'derivs': [0, 0]
}},
timing_units='secs',
interscan_interval=TR,
model_serial_correlations='FAST'),
name="level1design")
# EstimateModel - estimate the parameters of the model
level1estimate = Node(EstimateModel(estimation_method={'Classical': 1}),
name="level1estimate")
# EstimateContrast - estimates contrasts
level1conest = Node(EstimateContrast(), name="level1conest")
# Get Subject Info - get subject specific condition information
getsubjectinfo = Node(Function(input_names=['subject_id'],
output_names=['subject_info'],
function=subjectinfo_func),
name='getsubjectinfo')
# Infosource - a function free node to iterate over the list of subject names
infosource = Node(IdentityInterface(fields=['subject_id', 'contrasts'],
contrasts=contrast_list),
name="infosource")
infosource.iterables = [('subject_id', subject_list)]
# SelectFiles - to grab the data (alternativ to DataGrabber)
smooth_dir = opj(experiment_dir, 'smooth_nomask', 'preproc')
templates = {
'func': opj(smooth_dir, 'sub-{subject_id}', '*run-*_fwhm-8_bold.nii')
}
selectfiles = Node(SelectFiles(templates,
base_directory=experiment_dir,
sort_filelist=True),
name="selectfiles")
# Datasink - creates output folder for important outputs
datasink = Node(DataSink(base_directory=experiment_dir,
container=output_dir),
name="datasink")
# Use the following DataSink output substitutions
substitutions = [('_subject_id_', 'sub-')]
datasink.inputs.substitutions = substitutions
# Initiation of the 1st-level analysis workflow
l1analysis = Workflow(name='l1analysis')
l1analysis.base_dir = opj(experiment_dir, working_dir)
# Connect up the 1st-level analysis components
l1analysis.connect([
(infosource, selectfiles, [('subject_id', 'subject_id')]),
(infosource, getsubjectinfo, [('subject_id', 'subject_id')]),
(getsubjectinfo, modelspec, [('subject_info', 'subject_info')]),
(infosource, level1conest, [('contrasts', 'contrasts')]),
(selectfiles, modelspec, [('func', 'functional_runs')]),
(modelspec, level1design, [('session_info', 'session_info')]),
(level1design, level1estimate, [('spm_mat_file', 'spm_mat_file')]),
(level1estimate, level1conest, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image', 'residual_image')]),
(level1conest, datasink, [
('spm_mat_file', '1stLevel.@spm_mat'),
('spmT_images', '1stLevel.@T'),
('con_images', '1stLevel.@con'),
('spmF_images', '1stLevel.@F'),
('ess_images', '1stLevel.@ess'),
]),
])
return l1analysis
def second_level(contrast_list,
experiment_dir,
first_level_dir,
output_dir,
mask_path='/data/group_mask.nii.gz',
working_dir='workingdir'):
"""define second level model"""
# Gunzip - unzip the mask image
gunzip = Node(Gunzip(in_file=mask_path), name="gunzip")
# OneSampleTTestDesign - creates one sample T-Test Design
onesamplettestdes = Node(OneSampleTTestDesign(), name="onesampttestdes")
# EstimateModel - estimates the model
level2estimate = Node(EstimateModel(estimation_method={'Classical': 1}),
name="level2estimate")
# EstimateContrast - estimates group contrast
level2conestimate = Node(EstimateContrast(group_contrast=True),
name="level2conestimate")
cont1 = ['Group', 'T', ['mean'], [1]]
level2conestimate.inputs.contrasts = [cont1]
# Threshold - thresholds contrasts
level2thresh = Node(
Threshold(
contrast_index=1,
use_topo_fdr=True,
# use_fwe_correction=True,
use_fwe_correction=False,
extent_threshold=0,
height_threshold=0.001,
# height_threshold=0.05,
height_threshold_type='p-value',
extent_fdr_p_threshold=0.05),
name="level2thresh")
# Infosource - a function free node to iterate over the list of subject names
infosource = Node(IdentityInterface(fields=['contrast_id']),
name="infosource")
infosource.iterables = [('contrast_id', contrast_list)]
# SelectFiles - to grab the data (alternativ to DataGrabber)
firstlev_dir = opj(experiment_dir, first_level_dir, '1stLevel')
templates = {'cons': opj(firstlev_dir, 'sub-*', '{contrast_id}.nii')}
selectfiles = Node(SelectFiles(templates,
base_directory=experiment_dir,
sort_filelist=True),
name="selectfiles")
# Datasink - creates output folder for important outputs
datasink = Node(DataSink(base_directory=experiment_dir,
container=output_dir),
name="datasink")
# Use the following DataSink output substitutions
substitutions = [('_contrast_id_', '')]
datasink.inputs.substitutions = substitutions
# Initiation of the 2nd-level analysis workflow
l2analysis = Workflow(name='l2analysis')
l2analysis.base_dir = opj(experiment_dir, working_dir)
# Connect up the 2nd-level analysis components
l2analysis.connect([
(infosource, selectfiles, [('contrast_id', 'contrast_id')]),
(selectfiles, onesamplettestdes, [('cons', 'in_files')]),
(gunzip, onesamplettestdes, [('out_file', 'explicit_mask_file')]),
(onesamplettestdes, level2estimate, [('spm_mat_file', 'spm_mat_file')
]),
(level2estimate, level2conestimate, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image',
'residual_image')]),
(level2conestimate, level2thresh, [
('spm_mat_file', 'spm_mat_file'),
('spmT_images', 'stat_image'),
]),
(level2conestimate, datasink, [('spm_mat_file', '2ndLevel.@spm_mat'),
('spmT_images', '2ndLevel.@T'),
('con_images', '2ndLevel.@con')]),
(level2thresh, datasink, [('thresholded_map', '2ndLevel.@threshold')]),
])
return l2analysis
def prepare_oneT():
outdirs = [
'/home/peter/Desktop/prepare/rest/output/sec_level/one_sample/alff_analysis_noglob/',
'/home/peter/Desktop/prepare/rest/output/sec_level/one_sample/falff_analysis_noglob/'
]
for n, outdir in enumerate(outdirs):
stroke_hdir = '/home/peter/Desktop/prepare/rest/output/'
all_stroke_files = glob.glob(stroke_hdir + '/*/f_alff/' + maps[n])
stroke_files = [x for x in all_stroke_files if '12months' not in x]
stroke_files.sort()
#Convert to DataFrame, cull subs that aren't in preprocessed data
stroke_files = pd.DataFrame(stroke_files, columns=['files'])
if n == 0:
stroke_files['initials'] = os.path.split(
os.path.split(stroke_files['files'][n])[0])[0][-2:]
else:
stroke_files['initials'] = os.path.split(
os.path.split(stroke_files['files'][n])[0])[0][-2:]
cull_list = stroke_files['initials'].isin(df['initials'])
df_cull = df[cull_list]
df_cull.rename(columns={'Unnamed: 0': 'old_idx'}, inplace=True)
#Specially remove the second HK (15-28) who refused to die...
#df_cull = df_cull[df_cull['old_idx'] != 194]
#df_cull.reset_index(inplace = True, drop = True)
try:
np.all(df_cull['initials'] == stroke_files['initials']) == 1
stroke_files['madrs_3m'] = df_cull['MADRS_score_3mth']
except:
print 'Data are not of identical length. Error in script.'
cov = {
'vector': stroke_files['madrs_3m'].tolist(),
'name': 'madrs_3m',
'interaction': 1,
'centering': 1
}
ttest = Node(OneSampleTTestDesign(), name='OneSampleT')
ttest.inputs.in_files = stroke_files['files'].tolist()
ttest.inputs.covariates = cov
ttest.inputs.explicit_mask_file = mask_file
modelEst = Node(EstimateModel(), name='EstimateModel')
modelEst.inputs.estimation_method = {'Classical': 1}
conEst = Node(EstimateContrast(), name='EstimateContrasts')
con_1 = ('Stroke', 'T', ['mean', 'madrs_3m'], [1.0, 0.0])
con_2 = ('Covariate', 'T', ['mean', 'madrs_3m'], [0.0, 1.0])
contrasts = [con_1, con_2]
conEst.inputs.contrasts = contrasts
conEst.inputs.group_contrast = True
l2analysis = Workflow(name='l2analysis')
l2analysis.base_dir = outdir
l2analysis.connect([
(ttest, modelEst, [('spm_mat_file', 'spm_mat_file')]),
(modelEst, conEst, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image', 'residual_image')]),
])
l2analysis.write_graph(graph2use='colored')
l2analysis.run('MultiProc', plugin_args={'n_procs': 1})
def prepare_indT_madrs_int(stand):
outdirs = [
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/alff_analysis_noglob/',
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/falff_analysis_noglob/',
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/falff_slow_5_analysis_noglob/',
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/falff_slow_4_analysis_noglob/'
]
#outdirs = ['/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/alff_analysis/', '/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/falff_analysis/', '/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/falff_slow_5_analysis/', '/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_interaction/falff_slow_4_analysis/']
for n, outdir in enumerate(outdirs):
stroke_hdir = '/home/peter/Desktop/prepare/rest/output/'
all_stroke_files = glob.glob(stroke_hdir + '/*/f_alff/' + maps[n])
stroke_files = [x for x in all_stroke_files if '12months' not in x]
stroke_files.sort()
#Convert to DataFrame, cull subs that aren't in preprocessed data
stroke_files = pd.DataFrame(stroke_files, columns=['files'])
if n == 0:
stroke_files['initials'] = os.path.split(
os.path.split(stroke_files['files'][n])[0])[0][-2:]
else:
stroke_files['initials'] = os.path.split(
os.path.split(stroke_files['files'][n])[0])[0][-2:]
cull_list = stroke_files['initials'].isin(df['initials'])
df_cull = df[cull_list]
df_cull.rename(columns={'Unnamed: 0': 'old_idx'}, inplace=True)
#Specially remove the second HK (15-28) who refused to die...
#df_cull = df_cull[df_cull['old_idx'] != 194]
#df_cull.reset_index(inplace = True, drop = True)
try:
np.all(df_cull['initials'] == stroke_files['initials']) == 1
stroke_files['madrs_3m'] = df_cull['MADRS_score_3mth']
except:
print 'Data are not of identical length. Error in script.'
#ADD LOW / HIGH GROUPS TO CULLED DF
stroke_files['group'] = np.zeros(len(stroke_files))
stroke_files['group'][stroke_files['madrs_3m'] <= 8] = 1
stroke_files['group'][stroke_files['madrs_3m'] > 8] = 2
stroke_files.to_csv(
'/home/peter/Dropbox/post_doc/florey_leeanne/study_scripts/prepare/prepare_files_madrs_grouped.csv'
)
#SPLIT GROUPS & MAKE COVARIATE DEFINITIONS
stroke_low = stroke_files[stroke_files['madrs_3m'] <= 8]
stroke_high = stroke_files[stroke_files['madrs_3m'] > 8]
#CALC MADRS STANDARDIZATION (0 - 1)
lmin = stroke_low.madrs_3m.min()
lmax = stroke_low.madrs_3m.max()
stroke_low['madrs_3m_stand'] = (stroke_low.madrs_3m - lmin) / (lmax -
lmin)
hmin = stroke_high.madrs_3m.min()
hmax = stroke_high.madrs_3m.max()
stroke_high['madrs_3m_stand'] = (stroke_high.madrs_3m - hmin) / (hmax -
hmin)
if stand == 1:
outdir = outdir[:-1] + '_stand'
cov1 = np.zeros(len(stroke_files))
cov1[:len(stroke_low)] = stroke_low['madrs_3m_stand']
cov1 = {
'vector': cov1.tolist(),
'name': 'low_madrs',
'interaction': 1,
'centering': 5
}
cov2 = np.zeros(len(stroke_files))
cov2[len(stroke_low):] = stroke_high['madrs_3m_stand']
cov2 = {
'vector': cov2.tolist(),
'name': 'high_madrs',
'interaction': 1,
'centering': 5
}
elif stand == 0:
cov1 = np.zeros(len(stroke_files)).astype(int)
cov1[:len(stroke_low)] = stroke_low['madrs_3m'].astype(int)
cov1 = {
'vector': cov1.tolist(),
'name': 'low_madrs',
'interaction': 1,
'centering': 5
}
cov2 = np.zeros(len(stroke_files)).astype(int)
cov2[len(stroke_low):] = stroke_high['madrs_3m'].astype(int)
cov2 = {
'vector': cov2.tolist(),
'name': 'high_madrs',
'interaction': 1,
'centering': 5
}
else:
print(
'Invalid value %i. Please enter either 0 for non-standardised or 1 for standardised'
% (stand), )
#INITIALISE T-TEST OBJECT
ttest = Node(TwoSampleTTestDesign(), name='TwoSampleT')
ttest.inputs.group1_files = stroke_low['files'].tolist()
ttest.inputs.group2_files = stroke_high['files'].tolist()
ttest.inputs.covariates = [cov1, cov2]
#ttest.inputs.explicit_mask_file = mask_file
#Estimate model (aka betas)
modelEst = Node(EstimateModel(), name='EstimateModel')
modelEst.inputs.estimation_method = {'Classical': 1}
#Estimate contrasts
conEst = Node(EstimateContrast(), name='EstimateContrasts')
con_1 = ('all', 'T',
['Group_{1}', 'Group_{2}', 'low_madrs',
'high_madrs'], [1.0, 1.0, 0.0, 0.0])
con_2 = ('low', 'T',
['Group_{1}', 'Group_{2}', 'low_madrs',
'high_madrs'], [1.0, 0.0, 0.0, 0.0])
con_3 = ('high', 'T',
['Group_{1}', 'Group_{2}', 'low_madrs',
'high_madrs'], [0.0, 1.0, 0.0, 0.0])
con_4 = ('low GT high', 'T',
['Group_{1}', 'Group_{2}', 'low_madrs',
'high_madrs'], [1.0, -1.0, 0.0, 0.0])
con_5 = ('high GT low', 'T',
['Group_{1}', 'Group_{2}', 'low_madrs',
'high_madrs'], [-1.0, 1.0, 0.0, 0.0])
con_6 = ('low vs. high', 'T',
['Group_{1}', 'Group_{2}', 'low_madrs',
'high_madrs'], [0.0, 0.0, 1.0, -1.0])
con_7 = ('high vs. low', 'T',
['Group_{1}', 'Group_{2}', 'low_madrs',
'high_madrs'], [0.0, 0.0, -1.0, 1.0])
contrasts = [con_1, con_2, con_3, con_4, con_5, con_6, con_7]
conEst.inputs.contrasts = contrasts
conEst.inputs.group_contrast = True
#RUN
l2analysis = Workflow(name='l2analysis')
l2analysis.base_dir = outdir
l2analysis.connect([
(ttest, modelEst, [('spm_mat_file', 'spm_mat_file')]),
(modelEst, conEst, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image', 'residual_image')]),
])
l2analysis.write_graph(graph2use='colored')
l2analysis.run('MultiProc', plugin_args={'n_procs': 1})
def prepare_indT_madrs():
outdirs = [
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_main/alff_analysis_noglob/',
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_main/falff_analysis_noglob/',
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_main/falff_slow_5_analysis_noglob/',
'/home/peter/Desktop/prepare/rest/output/sec_level/two_sample_madrs_main/falff_slow_4_analysis_noglob/'
]
for n, outdir in enumerate(outdirs):
stroke_hdir = '/home/peter/Desktop/prepare/rest/output/'
all_stroke_files = glob.glob(stroke_hdir + '/*/f_alff/' + maps[n])
stroke_files = [x for x in all_stroke_files if '12months' not in x]
stroke_files.sort()
#Convert to DataFrame, cull subs that aren't in preprocessed data
stroke_files = pd.DataFrame(stroke_files, columns=['files'])
if n == 0:
stroke_files['initials'] = os.path.split(
os.path.split(stroke_files['files'][n])[0])[0][-2:]
else:
stroke_files['initials'] = os.path.split(
os.path.split(stroke_files['files'][n])[0])[0][-2:]
cull_list = stroke_files['initials'].isin(df['initials'])
df_cull = df[cull_list]
df_cull.rename(columns={'Unnamed: 0': 'old_idx'}, inplace=True)
#Specially remove the second HK (15-28) who refused to die...
#df_cull = df_cull[df_cull['old_idx'] != 194]
#df_cull.reset_index(inplace = True, drop = True)
try:
np.all(df_cull['initials'] == stroke_files['initials']) == 1
stroke_files['madrs_3m'] = df_cull['MADRS_score_3mth']
except:
print 'Data are not of identical length. Error in script.'
#SPLIT GROUPS & MAKE COVARIATE DEFINITIONS
stroke_low = stroke_files[stroke_files['madrs_3m'] <= 8]
stroke_high = stroke_files[stroke_files['madrs_3m'] > 8]
#INITIALISE T-TEST OBJECT
ttest = Node(TwoSampleTTestDesign(), name='TwoSampleT')
ttest.inputs.group1_files = stroke_low['files'].tolist()
ttest.inputs.group2_files = stroke_high['files'].tolist()
ttest.inputs.explicit_mask_file = mask_file
#Estimate model (aka betas)
modelEst = Node(EstimateModel(), name='EstimateModel')
modelEst.inputs.estimation_method = {'Classical': 1}
#Estimate contrasts
conEst = Node(EstimateContrast(), name='EstimateContrasts')
con_1 = ('mean_low', 'T', ['Group_{1}', 'Group_{2}'], [1.0, 0.0])
con_2 = ('mean_high', 'T', ['Group_{1}', 'Group_{2}'], [0.0, 1.0])
con_3 = ('Low>High', 'T', ['Group_{1}', 'Group_{2}'], [1.0, -1.0])
con_4 = ('High>Low', 'T', ['Group_{1}', 'Group_{2}'], [-1.0, 1.0])
contrasts = [con_1, con_2, con_3, con_4]
conEst.inputs.contrasts = contrasts
conEst.inputs.group_contrast = True
#RUN
l2analysis = Workflow(name='l2analysis')
l2analysis.base_dir = outdir
l2analysis.connect([
(ttest, modelEst, [('spm_mat_file', 'spm_mat_file')]),
(modelEst, conEst, [('spm_mat_file', 'spm_mat_file'),
('beta_images', 'beta_images'),
('residual_image', 'residual_image')]),
])
l2analysis.write_graph(graph2use='colored')
l2analysis.run('MultiProc', plugin_args={'n_procs': 1})
