Running groupBaseline twice results in incorrect summary stats

[Jason Vander Heiden]

Passing a Baseline object through groupBaseline() twice seems to mangle the summary stats and/or somehow confuses plotBaselineSummary(). More details when I figure out exactly what's wrong.

[Jason Vander Heiden]

norm=100 in calcBaselineSigma() is to blame. The second pass makes the sigma score 100-fold lower than it should be. I haven't tested it yet, but after skimming the code I suspect calcBaselinePvalue() suffers from the same problem.

[Jason Vander Heiden]

Fixed the sigma values in 913e3db. We still need to check the p-value part.

[Jason Vander Heiden]

Added normalization to groupBaseline() in 8db9855. I think this is fixed everywhere. Reopen if it pops up somewhere.