Reading data for use in Semi-blind machine learning (SML-EL)

The program vreadconnectome prepares inputs for the program vsml. The purpose of vreadconnectome is to convert the connectome data into the format expected by the program vsml. It is assumed that the connectome data has been prepared using some other software package.

The input into vreadconnectome is a text-file in csv format containing the comma-separated connectome of one subject. Since the connectome is symmetric, the connectome of one subject has K*(K-1)/2 entries, where K is the number of parcels/components of the atlas. E.g. if the K=200, then there should be 19900 entries in the connectome. Typically, such an entry is a number specifying the (tangent) correlation between two fMRI time series.

No comments (“#” or “%”) are allowed in the input file. Also, there should not be any blank spaces.

The output filename must have the ending “.v”. This is the convention for the vista-format.

Example:

vreadconnectome -in connectome_subj007.csv -ncomponents 200 -out connectome_subj007.v

Parameters of ‘vreadconnectome’:

-help

Prints usage information.

-in

Input file (csv format)

-out

Output file (vista-format)

-ncomponents

Number of components/parcels in the atlas/decomposition