Moment0 - Githubissues

GBTAmmoniaSurvey / GAS

observing scripts and files related to the GBT Ammonia Survey (GAS, PI: Jaime E Pineda & Rachel Friesen)

MIT License

6 stars 13 forks source link

Moment0 #119

Closed jpinedaf closed 8 years ago

jpinedaf commented 8 years ago

this is the initial steps to improve moment map calculation (Issue #73)

WIP, do not merge yet, but I'd like to hear your thoughts @keflavich @low-sky @rfriesen

List of issues at submission:

[x] checked integrated intensity map with first look map, and there is a clear difference in total normalization --> It arises from the first look pipeline calculating the integrated intensity only of the NH3(1,1) main component
[x] it is slow
[x] Is this really the most efficient way to use SpectralCube? --> It looks like that

keflavich commented 8 years ago

This looks fine. I would generally use ~1-sigma or maybe even 2-sigma as the threshold.

low-sky commented 8 years ago

This seems like a fine overall approach once the actual call from the pyspeckit dependency is worked out reliably. Calling the eps threshold will give a nice broad window. However, does this need some failover for when there is nothing in the fit cube or do we want to leave those as zero?

jpinedaf commented 8 years ago

maybe the default should be to use the channels around +-3 mean sigma_v centered at the mean velocity of the map

jpinedaf commented 8 years ago

new version now calculates the map for the NH3(2,2) line.

Better determination of integrated intensity if there is not detection

These are better looking maps, except for some offending pixels (large sigma_v)