The following analysis was performed with Lee+12 map: /d/bip3/ezbc/perseus/data/av/perseus_av_lee12_iris_regrid_planckres.fits
Before reorganizing the code, the MLE parameters for Perseus using the Lee+12 map were the following:
however this run did not allow for an intercept to vary. I am currently running the analysis allowing the intercept to vary to make a fair comparison with the subsequent tests.
Below are the likelihood spaces using the Lee+12 IRIS image. The map has not had a background subtracted from it.
The slightly negative intercept is curious. A mag background is present in the data.
Below is a run while setting the intercept to 0. This favors a slightly lower DGR and width.
I subtracted a background of mag from the Lee+12 map before running the MLE analysis. The resulting likelihood space are below. The velocity width and DGR are exactly the same as derived from the original map. However the intercept is around mag. We expect for this intercept to be around 0, because the background has been subtracted from it.
It seems as though the calculations are deriving the correct intercept, but then subtracting the intercept away in likelihood space…incorrectly recording the intercept.
Binned images / masking progression
Below are the original and binned maps for Perseus at each progression of calculating the MLE. The masks do not vary much between the iterations of changing the width to mask.