The module nipy.labs.utils.simul_multisubject_fmri_dataset contains a various functions to create simulated activation maps in two, three and four dimensions. A 2D example is surrogate_2d_dataset(). The functions can position various activations and add noise, both as background noise and jitter in the activation positions and amplitude.
These functions can be useful to test methods.
Create surrogate (simulated) 2D activation data with spatial noise
Parameters : | n_subj: integer, optionnal :
shape=(30,30): tuple of integers, :
sk: float, optionnal :
noise_level: float, optionnal :
pos: 2D ndarray of integers, optionnal :
ampli: 1D ndarray of floats, optionnal :
spatial_jitter: float, optionnal :
signal_jitter: float, optionnal :
width: float or ndarray, optionnal :
width_jitter: float :
out_text_file: string or None, optionnal :
out_image_file: string or None, optionnal :
seed=False: int, optionnal :
|
---|---|
Returns : | dataset: 3D ndarray :
|
Create surrogate (simulated) 3D activation data with spatial noise.
Parameters : | n_subj: integer, optionnal :
shape=(20,20,20): tuple of 3 integers, :
mask=None: Nifti1Image instance, :
sk: float, optionnal :
noise_level: float, optionnal :
pos: 2D ndarray of integers, optionnal :
ampli: 1D ndarray of floats, optionnal :
spatial_jitter: float, optionnal :
signal_jitter: float, optionnal :
width: float or ndarray, optionnal :
out_text_file: string or None, optionnal :
out_image_file: string or None, optionnal :
seed=False: int, optionnal :
|
---|---|
Returns : | dataset: 3D ndarray :
|
Create surrogate (simulated) 3D activation data with spatial noise.
Parameters : | shape = (20, 20, 20): tuple of integers, :
mask=None: brifti image instance, :
n_scans: int, optional, :
n_sess: int, optional, :
dmtx: array of shape(n_scans, n_rows), :
sk: float, optionnal :
noise_level: float, optionnal :
signal_level: float, optional, :
out_image_file: string or list of strings or None, optionnal :
seed=False: int, optionnal :
|
---|---|
Returns : | dataset: a list of n_sess ndarray of shape :
|