Make megaplot nicer

A bunch of megaplot updates.

• Aesthetic updates for pages run with --fullOnly
• Bayes factors for --fullOnly runs (ratio of iterations spent in full vs signal/glitch only)
• Fix model dimension histograms so bin is centered around integer
• Made megaplot compatible with an arbitrary number of detectors (used to be hard coded to three)
• fixed some deprecation warnings
• Got rid of a lot of useless code

Examples:

fullOnly

4 IFO

Hi @meg.millhouse the residual histograms don't seem to have been produced for the 4 IFO run you have linked. Did you perhaps use an older version of megaplot?

I thought the residual histograms didn't work if it was run without --bayesline?

Ah, didn't realize the run was without --bayesLine. I checked the code and it expects a post/{1}/{1}_median_PSD_{0}.dat style PSD file. But, I wonder if we can feed it the PSD for non bayesline cases.

merged

mentioned in commit 7e353af6