These are the counts produced by cty_sub_big.cc:
     
     total = 95412
     total_sub = 15000
     yes = 4843
     yes_sub = 2430
     no = 90569
     no_sub = 12570
     freq = 0.0507588
     freq_sub = 0.162
     OSR = 3.19156

The size of the raw data set is 95412 of which 15000 were randomly
sampled for cty_sub.dat.

The proportion of donors in the raw data set is 5.07588%.

The proportion of donors in the subset is 16.2%.

This is an over sampling rate of 0.162/0.0507588 = 3.19156