The format of the data file is:

k: in cardinality constraint;
w0: initial hold;
ri: return rate;
cov: covariance 

The first five are dataset from OR lib, the last one is derived from a real world problem.