Opened 10 years ago
Closed 10 years ago
#219 closed defect (fixed)
Erroneous behaviour in Negative-search mode
Reported by: | MatthewWhiting | Owned by: | MatthewWhiting |
---|---|---|---|
Priority: | normal | Milestone: | Release-1.6 |
Component: | Output | Version: | 1.5 |
Severity: | major | Keywords: | |
Cc: |
Description
Running Duchamp on the PKS1830 cube from BETA shows up a number of errors when running with flagNegative=true.
A lot seem to arise from doing the parameterisation on the cube after re-inversion. So the peak spectrum gets the spectrum with the least amount of absorption; the axis-finding fits to the positive pixels; the moment map plot doesn't show the interesting pixels; and the values for peak and integrated flux need looking at.
It might also be useful in flagNegative mode to give the average continuum level for the detection. This is particularly relevant if we fit a baseline - what is the average continuum level from which we get some absorption? This would allow a quick determination of peak optical depth :)
Change History (2)
comment:1 Changed 10 years ago by
Status: | new → assigned |
---|
comment:2 Changed 10 years ago by
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
Think I might close this ticket, as this is probably sufficient for the next release. Want to add something to the User Guide, but this can be a new ticket.
Also, add a new ticket for (maybe) next release to improve the parameterisation of these sources.
Changeset [1361] largely solves the main problem identified in the ticket.
The problem is particularly apparent when running negative source searches with baseline subtraction. The sources identified are below the baseline but still positive.
I've had to put specific switches in the calls to invert and replaceBaseline to handle the different situations. Now the flow is:
I've also removed a couple of additional calls to calcFluxes that were being made at plotting time. This will only help speed things up a bit!