Empty flux.fits file in certain model sets

[mattpovich]

Hi Tom,

I have been convolving the new YSO models with UKIRT JHK filters since these aren't provided with the model packages (if anyone needs them, they can ask me for them plus the script I used to generate them, calling your sedfitter packages).

I'm using 6 of your model sets. sp--s-i, sp--h-i, spu-smi, and spu-hmi have all been working just fine. But when I try to convolve the spubsmi or spubhmi models with new filters, it crashes. I traced this back to empty flux.fits files in the tarballs I downloaded from your zenodo repository.

Thanks,

Matt

I had a couple of problems trying to unpack spubsmi.tar and spubhmi.tar that I foolishly ignored the first time around.

Using MacOS's built-in archive utility I got a "permission denied" error at the end, and the archive wasn't expanded.

Below is what happens when I tried to expand one of them from the command-line:

% tar xvzf spubsmi.tar
x spubsmi/
x spubsmi/stellar.fits
x spubsmi/flux.fits: Attempt to write to an empty file
x spubsmi/parameters.fits
x spubsmi/models.conf
x spubsmi/convolved/
x spubsmi/convolved/BV.fits
x spubsmi/convolved/M2.fits
x spubsmi/convolved/SPIR2.fits
x spubsmi/convolved/M1.fits
x spubsmi/convolved/BU.fits
x spubsmi/convolved/SPIR1.fits
x spubsmi/convolved/W2.fits
x spubsmi/convolved/M3.fits
x spubsmi/convolved/2J.fits
x spubsmi/convolved/PACS2.fits
x spubsmi/convolved/PACS3.fits
x spubsmi/convolved/W1.fits
x spubsmi/convolved/I1.fits
x spubsmi/convolved/WISE4.fits
x spubsmi/convolved/WISE1.fits
x spubsmi/convolved/BI.fits
x spubsmi/convolved/BR.fits
x spubsmi/convolved/I4.fits
x spubsmi/convolved/I2.fits
x spubsmi/convolved/WISE3.fits
x spubsmi/convolved/PACS1.fits
x spubsmi/convolved/2H.fits
x spubsmi/convolved/I3.fits
x spubsmi/convolved/WISE2.fits
x spubsmi/convolved/2K.fits
x spubsmi/convolved/SPIR3.fits
x spubsmi/convolved/BB.fits
tar: Error exit delayed from previous errors.

[mpound]

Hi Matt,

It looks like the tar file was corrupted on download or you ran out of space? I have not had any encountered any empty flux.fits files.

I'd like to see your UKIRT script and filter input data. I have convolved SDSS filters, but in testing against a just a star (s---s-i.tar), the fits are wonky. A working example successful convolution (and testing thereof) would help. I'd be happy to share back once I get it working.

Marc

[mattpovich]

Hi Marc, My script is really simple, not even a script really, just a short series of commands in the attached file. Note that I've hard-coded the specific paths to the various YSO model directories on my own system, as well as the particular central wavelengths of the UKIRT JHK filters (in µm). I've not tested this on the kurucz stellar atmosphere models (still have Tom's old UKIRT convolutions in my directory for those). But so far the fits have worked just fine on all the Robitaille (2017) models I've tried. Matt

…

On Wed, Aug 8, 2018 at 2:47 PM Marc Pound ***@***.***> wrote: Hi Matt, It looks like the tar file was corrupted on download or you ran out of space? I have not had any encountered any empty flux.fits files. I'd like to see your UKIRT script and filter input data. I have convolved SDSS filters, but in testing against a just a star (s---s-i.tar), the fits are wonky. A working example successful convolution (and testing thereof) would help. I'd be happy to share back once I get it working. Marc — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#3 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AnrfICbMoif-OFDg9nOIG4ZtRNNJcXiYks5uO1xWgaJpZM4Vpchh> .

-- --

________________________________ Matthew S. Povich Associate Professor Department of Physics and Astronomy College of Science Building 8, Rm 216 Cal Poly Pomona http://www.cpp.edu/~mspovich/ mspovich@cpp.edu Office: (909) 869-3608 Mobile: (617) 306-0411

[mpound]

Hi Matt,

I got mine working. My problem was the frequencies in the data files were in the wrong order. This causes convolve to silently make all values zero because Filter expects frequencies in a certain order.
thanks,

I now have GAIA, SDSS, SOFIA FORCAST and SOFIA HAWC.

Marc

[mattpovich]

Yes, I had that problem initially too!

On Fri, Aug 24, 2018 at 1:23 PM Marc Pound ***@***.***> wrote: Hi Matt, I got mine working. My problem was the frequencies in the data files were in the wrong order. This causes convolve to silently make all values zero because Filter expects frequencies in a certain order. thanks, Marc — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#3 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AnrfIM1Xboq0DDMbaFUxAmXpM59O18HRks5uUGCtgaJpZM4Vpchh> .

-- Sent from Gmail Mobile

[jsmonzon]

hey @mattpovich,

I have encountered the same problem with empty flux.fits files on the same two downloads:
spubsmi.tar and spubhmi.tar

Any tips? Did you end up just not using these models?
Thanks!

JSM