This project is read-only.

Sparse File Format

Files such as PhyloD's predictorFile and targetFile are in the sprase file format.

This format is three-column and tab-delimited text file with headers var (variable), cid (case identifier), and val (value). For example, suppose that AnHla is the name of an HLA.

This HLA file starts:

var cid val
AnHla 1 1
AnHla 2 0
AnHla 3 0
AnHla 4 1
AnHla 5 1
... ... ...

This says that the patient with cid 1 has HLA "AnHla". Patients 2 and 3 don't. Patients 4 and 5 do.

Missing data is supported. To say, for example, that you don't know if patient 4 has HLA "AnHLA", just leave out the "AnHla 4" line from the file.

The values in the var column may be any string. The values in the cid column may be any string. The values in the val column should be either 0 or 1.

Last edited Apr 4, 2007 at 7:01 PM by CarlK, version 9


No comments yet.