maxen running slowly

Is MIRIAD being a pain? Let us know your experience.

Moderator: Mark.Wieringa

Post Reply
ycendes
Posts: 4
Joined: Wed Jan 11, 2017 8:15 am

maxen running slowly

Post by ycendes »

Hi,

As requested, I am posting to confirm that I'm having issues running maxen at a fast speed- one iteration takes at least a second. (mfcal runs great though, thanks!)

Details:

Command I am using: "maxen map=sn1987a.1.imap beam=sn1987a.1.beam out=sn1987a.1.icmp niters=200 rms=1e-7 region=@first.region"

As for the files I'm runing, they are available in the dropbox folder here.

Thanks again!
Mark.Wieringa
ATCA Expert
Posts: 297
Joined: Mon Feb 08, 2010 1:37 pm

Re: maxen running slowly

Post by Mark.Wieringa »

Hi,

Two things:
- a maxen iteration is not like a clean iteration, you'll only need 20-100 maxen iterations to deconvolve an image, whereas you might need several 1000's of clean iterations.

- I think maxen is not being very smart with the very large beam you're giving it. If you decrease the beam size to 2048 pixels or less it runs much faster (imsub in=beam out=beam.small "region=perc(20)"). Since your region is pretty small, you get the same results. The beam really only needs to be twice the size of the region to deconvolve. The region probably needs to be a bit larger than 'first.region' to avoid the edge effects.

Cheers,

Mark
ycendes
Posts: 4
Joined: Wed Jan 11, 2017 8:15 am

Re: maxen running slowly

Post by ycendes »

Ah ok- mea culpa, I didn't specify the image size correctly in the earlier invert command. Makes all the difference, maxen runs much faster now.

Thanks for the help and for the other tips! :)
Post Reply