Closed will-henney closed 2 years ago
I forgot to mention that another change is to not do any merging of the small-separation bins for Lagoon. It does not really help for this source. It is fine to have one bin that deviates from the fit:
Also, I left all the weights the same, since this seemed to help with the fit
The main result here is to fix the Lagoon strucfunc model fit. The reason that this was not working previously is that the lower limit on
sig2
was too high. This is now fixed along with other minor issues:bplot.STYLE
dict to allow tweaks to figure (e.g. label positions)CI*.ipynb
)