Date: Nov 27, 2012 3:16 PM
Author: Halitsky
Subject: Holy Cow! Look at your "average a1" slope regressed on Len Int

If you have a moment, regress your "average a1" on LenInt:


LenInt "average a1"
1 -0.487393513
2 -0.178292604
3 -0.164174277
4 -0.088771216
5 -0.104084177
6 -0.043135845
7 0.02423492
8 -0.021704118
9 0.097815884
10 -0.03499833
11 -0.01778355
12 -0.013260234


These data are for a1_N_1_S (fold a1, nonrandom method, dicodon set 1,
dicodon subset S),
and underlying regression is c on (u', u'^2), where u' = u/(1+u).

I have to believe that you're picking up on something quite strongly
here, but of course,
will have to wait to see what happens when all folds, sets, subsets
are done.