Fork me on GitHub

pyhrf.ui.jde module

class pyhrf.ui.jde.JDEAnalyser(outputPrefix='jde_', pass_error=True)

Bases: pyhrf.ui.analyser_ui.FMRIAnalyser

get_label()
class pyhrf.ui.jde.JDEMCMCAnalyser(sampler=<pyhrf.jde.models.BOLDGibbsSampler object>, osfMax=4, dtMin=0.4, dt=0.6, driftParam=4, driftType='polynomial', outputPrefix='jde_mcmc_', randomSeed=None, pass_error=True, copy_sampler=True)

Bases: pyhrf.ui.jde.JDEAnalyser

Class that wraps a JDE Gibbs Sampler to launch an fMRI analysis TODO: remove parameters about dt and osf (should go in HRF Sampler class), drift (should go in Drift Sampler class)

P_DRIFT_LFD_PARAM = 'driftParam'
P_DRIFT_LFD_TYPE = 'driftType'
P_DT = 'dt'
P_DTMIN = 'dtMin'
P_OSFMAX = 'osfMax'
P_RANDOM_SEED = 'randomSeed'
P_SAMPLER = 'sampler'
analyse_roi(atomData)

Launch the JDE Gibbs Sampler on a parcel-specific data set atomData :param - atomData: parcel-specific data :type - atomData: pyhrf.core.FmriData

Returns:JDE sampler object
enable_draft_testing()
packSamplerInput(roiData)
parametersComments = {'driftParam': 'Parameter of the drift modelling.\nIf drift is "polynomial" then this is the order of the polynom.\nIf drift is "cosine" then this is the cut-off period in second.', 'driftType': 'Either "cosine" or "polynomial" or "None"', 'dt': "If different from 0 or None:\nactual time resolution for the oversampled estimated signal (dtMin is ignored).\n Better when it's a multiple of the time of repetition", 'dtMin': 'Minimum time resolution for the oversampled estimated signal', 'sampler': 'Set of parameters for the sampling scheme'}
parametersToShow = ['dtMin', 'dt', 'driftType', 'driftParam', 'sampler']
pyhrf.ui.jde.jde_analyse(data=None, nbIterations=3, hrfModel='estimated', hrfNorm=1.0, hrfTrick=False, sampleHrfVar=True, hrfVar=1e-05, keepSamples=False, samplesHistPace=1)
pyhrf.ui.jde.runEstimationBetaEstim(params)
pyhrf.ui.jde.runEstimationSupervised(params)