.. AUTO-GENERATED FILE -- DO NOT EDIT!

nipype.interfaces.spm.model
===========================


:class:`EstimateContrast`
-------------------------


use spm_contrasts to estimate contrasts of interest

Examples
~~~~~~~~
>>> import nipype.interfaces.spm as spm
>>> est = spm.EstimateContrast()
>>> est.inputs.spm_mat_file = 'SPM.mat'
>>> cont1 = ('Task>Baseline','T', ['Task-Odd','Task-Even'],[0.5,0.5])
>>> cont2 = ('Task-Odd>Task-Even','T', ['Task-Odd','Task-Even'],[1,-1])
>>> contrasts = [cont1,cont2]
>>> est.inputs.contrasts = contrasts
>>> est.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	beta_images : (an existing file name)
		Parameter estimates of the design matrix
	contrasts : (a list of items which are a tuple of the form: (a string, 'T', a list of items which are a string, a list of items which are a float) or a tuple of the form: (a string, 'T', a list of items which are a string, a list of items which are a float, a list of items which are a float) or a tuple of the form: (a string, 'F', a list of items which are a tuple of the form: (a string, 'T', a list of items which are a string, a list of items which are a float) or a tuple of the form: (a string, 'T', a list of items which are a string, a list of items which are a float, a list of items which are a float)))
		List of contrasts with each contrast being a list of the form:
            [('name', 'stat', [condition list], [weight list], [session list])]. if
            session list is None or not provided, all sessions are used. For F
            contrasts, the condition list should contain previously defined
            T-contrasts.
	residual_image : (an existing file name)
		Mean-squared image of the residuals
	spm_mat_file : (an existing file name)
		Absolute path to SPM.mat

	[Optional]
	group_contrast : (a boolean)
		higher level contrast
		exclusive: use_derivs
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	paths : (a directory name)
		Paths to add to matlabpath
	use_derivs : (a boolean)
		use derivatives for estimation
		exclusive: group_contrast
	use_mcr : (a boolean)
		Run m-code using SPM MCR


Outputs:: 

	con_images : (an existing file name)
		contrast images from a t-contrast
	ess_images : (an existing file name)
		contrast images from an F-contrast
	spmF_images : (an existing file name)
		stat images from an F-contrast
	spmT_images : (an existing file name)
		stat images from a t-contrast
	spm_mat_file : (a file name)
		Updated SPM mat file

:class:`EstimateModel`
----------------------


Use spm_spm to estimate the parameters of a model

http://www.fil.ion.ucl.ac.uk/spm/doc/manual.pdf#page=71

Examples
~~~~~~~~
>>> est = EstimateModel()
>>> est.inputs.spm_mat_file = 'SPM.mat'
>>> est.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	estimation_method : (a dictionary with keys which are 'Classical' or 'Bayesian2' or 'Bayesian' and with values which are any value)
		Classical, Bayesian2, Bayesian (dict)
	spm_mat_file : (an existing file name)
		absolute path to SPM.mat

	[Optional]
	flags : (a string)
		optional arguments (opt)
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	paths : (a directory name)
		Paths to add to matlabpath
	use_mcr : (a boolean)
		Run m-code using SPM MCR


Outputs:: 

	RPVimage : (an existing file name)
		Resels per voxel image
	beta_images : (an existing file name)
		design parameter estimates
	mask_image : (an existing file name)
		binary mask to constrain estimation
	residual_image : (an existing file name)
		Mean-squared image of the residuals
	spm_mat_file : (a file name)
		Updated SPM mat file

:class:`FactorialDesign`
------------------------


Base class for factorial designs

http://www.fil.ion.ucl.ac.uk/spm/doc/manual.pdf#page=79

Inputs:: 

	[Optional]
	covariates : (a dictionary with keys which are 'vector' or 'name' or 'interaction' or 'centering' and with values which are any value)
		covariate dictionary {vector, name, interaction, centering}
	explicit_mask_file : (a file name)
		use an implicit mask file to threshold
	global_calc_mean : (a boolean)
		use mean for global calculation
		exclusive: global_calc_omit,global_calc_values
	global_calc_omit : (a boolean)
		omit global calculation
		exclusive: global_calc_mean,global_calc_values
	global_calc_values : (a list of items which are a float)
		omit global calculation
		exclusive: global_calc_mean,global_calc_omit
	global_normalization : (1 or 2 or 3)
		global normalization None-1, Proportional-2, ANCOVA-3
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	no_grand_mean_scaling : (a boolean)
		do not perform grand mean scaling
	paths : (a directory name)
		Paths to add to matlabpath
	spm_mat_dir : (an existing directory name)
		directory to store SPM.mat file (opt)
	threshold_mask_absolute : (a float)
		threshold using a proportion of the global value
		exclusive: threshold_mask_absolute,threshold_mask_none
	threshold_mask_none : (a boolean)
		do not use threshold masking
		exclusive: threshold_mask_absolute,threshold_mask_relative
	use_implicit_threshold : (a boolean)
		use implicit mask NaNs or zeros to threshold
	use_mcr : (a boolean)
		Run m-code using SPM MCR


Outputs:: 

	spm_mat_file : (an existing file name)
		SPM mat file

:class:`Level1Design`
---------------------


Generate an SPM design matrix

http://www.fil.ion.ucl.ac.uk/spm/doc/manual.pdf#page=61

Examples
~~~~~~~~

>>> level1design = Level1Design()
>>> level1design.inputs.timing_units = 'secs'
>>> level1design.inputs.interscan_interval = 2.5
>>> level1design.inputs.bases = {'hrf':{'derivs': [0,0]}}
>>> level1design.inputs.session_info = 'session_info.npz'
>>> level1design.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	bases : (a dictionary with keys which are 'hrf' or 'fourier' or 'fourier_han' or 'gamma' or 'fir' and with values which are any value)
		
            dict {'name':{'basesparam1':val,...}}
            name : string
                Name of basis function (hrf, fourier, fourier_han,
                gamma, fir)
                
                hrf :
                    derivs : 2-element list
                        Model  HRF  Derivatives. No derivatives: [0,0],
                        Time derivatives : [1,0], Time and Dispersion
                        derivatives: [1,1]
                fourier, fourier_han, gamma, fir:
                    length : int
                        Post-stimulus window length (in seconds)
                    order : int
                        Number of basis functions

	interscan_interval : (a float)
		Interscan interval in secs
	session_info	Session specific information generated by ``modelgen.SpecifyModel``
	timing_units : ('secs' or 'scans')
		units for specification of onsets

	[Optional]
	factor_info : (a list of items which are a dictionary with keys which are 'name' or 'levels' and with values which are any value)
		Factor specific information file (opt)
	global_intensity_normalization : ('none' or 'scaling')
		Global intensity normalization - scaling or none (opt)
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	mask_image : (an existing file name)
		Image  for  explicitly  masking the analysis (opt)
	mask_threshold : ('-Inf' or a float)
		Thresholding for the mask (opt, '-Inf')
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	microtime_onset : (a float)
		The onset/time-bin in seconds for alignment (opt)
	microtime_resolution : (an integer)
		Number of time-bins per scan in secs (opt)
	model_serial_correlations : ('AR(1)' or 'none')
		Model serial correlations AR(1) or none (opt)
	paths : (a directory name)
		Paths to add to matlabpath
	spm_mat_dir : (an existing directory name)
		directory to store SPM.mat file (opt)
	use_mcr : (a boolean)
		Run m-code using SPM MCR
	volterra_expansion_order : (1 or 2)
		Model interactions - yes:1, no:2 (opt)


Outputs:: 

	spm_mat_file : (an existing file name)
		SPM mat file

:class:`MultipleRegressionDesign`
---------------------------------


Create SPM design for multiple regression

Examples
~~~~~~~~

>>> mreg = MultipleRegressionDesign()
>>> mreg.inputs.in_files = ['cont1.nii','cont2.nii']
>>> mreg.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	in_files : (a list of at least 2 items which are an existing file name)
		List of files

	[Optional]
	covariates : (a dictionary with keys which are 'vector' or 'name' or 'interaction' or 'centering' and with values which are any value)
		covariate dictionary {vector, name, interaction, centering}
	explicit_mask_file : (a file name)
		use an implicit mask file to threshold
	global_calc_mean : (a boolean)
		use mean for global calculation
		exclusive: global_calc_omit,global_calc_values
	global_calc_omit : (a boolean)
		omit global calculation
		exclusive: global_calc_mean,global_calc_values
	global_calc_values : (a list of items which are a float)
		omit global calculation
		exclusive: global_calc_mean,global_calc_omit
	global_normalization : (1 or 2 or 3)
		global normalization None-1, Proportional-2, ANCOVA-3
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	include_intercept : (a boolean)
		Include intercept in design
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	no_grand_mean_scaling : (a boolean)
		do not perform grand mean scaling
	paths : (a directory name)
		Paths to add to matlabpath
	spm_mat_dir : (an existing directory name)
		directory to store SPM.mat file (opt)
	threshold_mask_absolute : (a float)
		threshold using a proportion of the global value
		exclusive: threshold_mask_absolute,threshold_mask_none
	threshold_mask_none : (a boolean)
		do not use threshold masking
		exclusive: threshold_mask_absolute,threshold_mask_relative
	use_implicit_threshold : (a boolean)
		use implicit mask NaNs or zeros to threshold
	use_mcr : (a boolean)
		Run m-code using SPM MCR
	user_covariates : (a dictionary with keys which are 'vector' or 'name' or 'centering' and with values which are any value)
		covariate dictionary {vector, name, centering}


Outputs:: 

	spm_mat_file : (an existing file name)
		SPM mat file

:class:`OneSampleTTestDesign`
-----------------------------


Create SPM design for one sample t-test

Examples
~~~~~~~~

>>> ttest = OneSampleTTestDesign()
>>> ttest.inputs.in_files = ['cont1.nii', 'cont2.nii']
>>> ttest.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	in_files : (a list of at least 2 items which are an existing file name)
		input files

	[Optional]
	covariates : (a dictionary with keys which are 'vector' or 'name' or 'interaction' or 'centering' and with values which are any value)
		covariate dictionary {vector, name, interaction, centering}
	explicit_mask_file : (a file name)
		use an implicit mask file to threshold
	global_calc_mean : (a boolean)
		use mean for global calculation
		exclusive: global_calc_omit,global_calc_values
	global_calc_omit : (a boolean)
		omit global calculation
		exclusive: global_calc_mean,global_calc_values
	global_calc_values : (a list of items which are a float)
		omit global calculation
		exclusive: global_calc_mean,global_calc_omit
	global_normalization : (1 or 2 or 3)
		global normalization None-1, Proportional-2, ANCOVA-3
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	no_grand_mean_scaling : (a boolean)
		do not perform grand mean scaling
	paths : (a directory name)
		Paths to add to matlabpath
	spm_mat_dir : (an existing directory name)
		directory to store SPM.mat file (opt)
	threshold_mask_absolute : (a float)
		threshold using a proportion of the global value
		exclusive: threshold_mask_absolute,threshold_mask_none
	threshold_mask_none : (a boolean)
		do not use threshold masking
		exclusive: threshold_mask_absolute,threshold_mask_relative
	use_implicit_threshold : (a boolean)
		use implicit mask NaNs or zeros to threshold
	use_mcr : (a boolean)
		Run m-code using SPM MCR


Outputs:: 

	spm_mat_file : (an existing file name)
		SPM mat file

:class:`PairedTTestDesign`
--------------------------


Create SPM design for paired t-test

Examples
~~~~~~~~

>>> pttest = PairedTTestDesign()
>>> pttest.inputs.paired_files = [['cont1.nii','cont1a.nii'],['cont2.nii','cont2a.nii']]
>>> pttest.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	paired_files : (a list of at least 2 items which are a list of from 2 to 2 items which are an existing file name)
		List of paired files

	[Optional]
	ancova : (a boolean)
		Specify ancova-by-factor regressors
	covariates : (a dictionary with keys which are 'vector' or 'name' or 'interaction' or 'centering' and with values which are any value)
		covariate dictionary {vector, name, interaction, centering}
	explicit_mask_file : (a file name)
		use an implicit mask file to threshold
	global_calc_mean : (a boolean)
		use mean for global calculation
		exclusive: global_calc_omit,global_calc_values
	global_calc_omit : (a boolean)
		omit global calculation
		exclusive: global_calc_mean,global_calc_values
	global_calc_values : (a list of items which are a float)
		omit global calculation
		exclusive: global_calc_mean,global_calc_omit
	global_normalization : (1 or 2 or 3)
		global normalization None-1, Proportional-2, ANCOVA-3
	grand_mean_scaling : (a boolean)
		Perform grand mean scaling
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	no_grand_mean_scaling : (a boolean)
		do not perform grand mean scaling
	paths : (a directory name)
		Paths to add to matlabpath
	spm_mat_dir : (an existing directory name)
		directory to store SPM.mat file (opt)
	threshold_mask_absolute : (a float)
		threshold using a proportion of the global value
		exclusive: threshold_mask_absolute,threshold_mask_none
	threshold_mask_none : (a boolean)
		do not use threshold masking
		exclusive: threshold_mask_absolute,threshold_mask_relative
	use_implicit_threshold : (a boolean)
		use implicit mask NaNs or zeros to threshold
	use_mcr : (a boolean)
		Run m-code using SPM MCR


Outputs:: 

	spm_mat_file : (an existing file name)
		SPM mat file

:class:`Threshold`
------------------


Topological FDR thresholding based on cluster extent/size. Smoothness is
estimated from GLM residuals but is assumed to be the same for all of the
voxels.

Examples
~~~~~~~~

>>> thresh = Threshold()
>>> thresh.inputs.spm_mat_file = 'SPM.mat'
>>> thresh.inputs.stat_image = 'spmT_0001.img'
>>> thresh.inputs.contrast_index = 1
>>> thresh.inputs.extent_fdr_p_threshold = 0.05
>>> thresh.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	contrast_index : (an integer)
		which contrast in the SPM.mat to use
	spm_mat_file : (an existing file name)
		absolute path to SPM.mat
	stat_image : (an existing file name)
		stat image

	[Optional]
	extent_fdr_p_threshold : (a float)
		p threshold on FDR corrected cluster size probabilities
	extent_threshold : (an integer)
		Minimum cluster size in voxels
	height_threshold : (a float)
		p-value for initial thresholding (defining clusters)
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	paths : (a directory name)
		Paths to add to matlabpath
	use_fwe_correction : (a boolean)
		whether to use FWE (Bonferroni) correction for initial threshold
	use_mcr : (a boolean)
		Run m-code using SPM MCR
	use_topo_fdr : (a boolean)
		whether to use FDR over cluster extent probabilities


Outputs:: 

	pre_topo_fdr_map : (an existing file name)
		Unknown
	thresholded_map : (an existing file name)
		Unknown

:class:`ThresholdStatistics`
----------------------------


Given height and cluster size threshold calculate theoretical probabilities
concerning false positives

Examples
~~~~~~~~

>>> thresh = ThresholdStatistics()
>>> thresh.inputs.spm_mat_file = 'SPM.mat'
>>> thresh.inputs.stat_image = 'spmT_0001.img'
>>> thresh.inputs.contrast_index = 1
>>> thresh.inputs.height_threshold = 4.56
>>> thresh.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	contrast_index : (an integer)
		which contrast in the SPM.mat to use
	height_threshold : (a float)
		stat value for initial thresholding (defining clusters)
	spm_mat_file : (an existing file name)
		absolute path to SPM.mat
	stat_image : (an existing file name)
		stat image

	[Optional]
	extent_threshold : (an integer)
		Minimum cluster size in voxels
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	paths : (a directory name)
		Paths to add to matlabpath
	use_mcr : (a boolean)
		Run m-code using SPM MCR


Outputs:: 

	clusterwise_P_FDR : (a float)
		Unknown
	clusterwise_P_RF : (a float)
		Unknown
	voxelwise_P_Bonf : (a float)
		Unknown
	voxelwise_P_FDR : (a float)
		Unknown
	voxelwise_P_RF : (a float)
		Unknown
	voxelwise_P_uncor : (a float)
		Unknown

:class:`TwoSampleTTestDesign`
-----------------------------


Create SPM design for two sample t-test

Examples
~~~~~~~~

>>> ttest = TwoSampleTTestDesign()
>>> ttest.inputs.group1_files = ['cont1.nii', 'cont2.nii']
>>> ttest.inputs.group2_files = ['cont1a.nii', 'cont2a.nii']
>>> ttest.run() # doctest: +SKIP

Inputs:: 

	[Mandatory]
	group1_files : (a list of at least 2 items which are an existing file name)
		Group 1 input files
	group2_files : (a list of at least 2 items which are an existing file name)
		Group 2 input files

	[Optional]
	covariates : (a dictionary with keys which are 'vector' or 'name' or 'interaction' or 'centering' and with values which are any value)
		covariate dictionary {vector, name, interaction, centering}
	dependent : (a boolean)
		Are the measurements dependent between levels
	explicit_mask_file : (a file name)
		use an implicit mask file to threshold
	global_calc_mean : (a boolean)
		use mean for global calculation
		exclusive: global_calc_omit,global_calc_values
	global_calc_omit : (a boolean)
		omit global calculation
		exclusive: global_calc_mean,global_calc_values
	global_calc_values : (a list of items which are a float)
		omit global calculation
		exclusive: global_calc_mean,global_calc_omit
	global_normalization : (1 or 2 or 3)
		global normalization None-1, Proportional-2, ANCOVA-3
	ignore_exception : (a boolean)
		Print an error message instead of throwing an exception in case the interface fails to run
	matlab_cmd : (a string)
		matlab command to use
	mfile : (a boolean)
		Run m-code using m-file
	no_grand_mean_scaling : (a boolean)
		do not perform grand mean scaling
	paths : (a directory name)
		Paths to add to matlabpath
	spm_mat_dir : (an existing directory name)
		directory to store SPM.mat file (opt)
	threshold_mask_absolute : (a float)
		threshold using a proportion of the global value
		exclusive: threshold_mask_absolute,threshold_mask_none
	threshold_mask_none : (a boolean)
		do not use threshold masking
		exclusive: threshold_mask_absolute,threshold_mask_relative
	unequal_variance : (a boolean)
		Are the variances equal or unequal between groups
	use_implicit_threshold : (a boolean)
		use implicit mask NaNs or zeros to threshold
	use_mcr : (a boolean)
		Run m-code using SPM MCR


Outputs:: 

	spm_mat_file : (an existing file name)
		SPM mat file
