[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: R: command: astsubtractsky and astnoisechisel
From: |
Mohammad Akhlaghi |
Subject: |
Re: R: command: astsubtractsky and astnoisechisel |
Date: |
Mon, 15 May 2023 20:08:17 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.11.0 |
Thanks Gabriele,
From your email, the main error message is in [1] below.
If you run the following command in a terminal (as the error message
suggests) you will see the list of HDUs in 'kernel.fits':
astfits /media/sf_Tesi/DatiInput/HighStellarDensity/NISP_Images/kernel.fits
You will see that this file only has a single HDU.
The default '--hdu=0' option that is suggested in the error message is
for the main input ('input_file' in your code). To specify the HDU for
reading the kernel, you should use '--khdu', it is described in this
part of the book:
https://www.gnu.org/software/gnuastro/manual/html_node/NoiseChisel-input.html
So if you add '--khdu=0' to your run of NoiseChisel, this error should
be fixed and you won't face a problem about the zero values on the edges
(with '--hdu=3', you were changing the main input's HDU!).
It is great that you are comparing NoiseChisel! Just be careful that
both NoiseChisel and SExtractor (as in any science software!) are highly
customizable and it is advised to read their manuals to understand your
results and optimize them for your science scenario. In the case of
NoiseChisel after doing the tutorial of the previous email, read this
tutorial to
https://www.gnu.org/software/gnuastro/manual/html_node/Detecting-large-extended-targets.html
Generally, it is easier to first test your commands on the command-line,
and then put them in a script ;-).
Another small tip: With Gnuastro's MakeProfiles command, you can easily
create any Gaussian kernel image. For example the output of the command
above is a 2D Gaussian profile with a FWHM of 2 pixels and trucated at 5
times the FWHM.
astmkprof --kernel=gaussian,2,5 --oversample=1 --output=kernel.fits
For more on this option, see:
https://www.gnu.org/software/gnuastro/manual//html_node/MakeProfiles-output-dataset.html
Cheers,
Mohammad
[1] Main error message
```
astnoisechisel:
/media/sf_Tesi/DatiInput/HighStellarDensity/NISP_Images/kernel.fits:
only has one HDU.
You should tell Gnuastro's command-line programs to look for data in the
primary HDU with the '--hdu=0' option (or '-h0'). For library users, you
can put "0" (a string literal) for the function's HDU argument. For
more, see the FOOTNOTE below.
Pro TIP: if your desired HDU has a name (value to 'EXTNAME' keyword), it
is best to just use that name with '--hdu' instead of relying on a
counter. You can see the list of HDUs in a FITS file (with their data
format, type, size and possibly HDU name) using Gnuastro's 'astfits'
program, for example:
astfits
/media/sf_Tesi/DatiInput/HighStellarDensity/NISP_Images/kernel.fits
FOOTNOTE -- When writing a new FITS file, Gnuastro leaves the pimary HDU
only for metadata. The output datasets (tables, images or cubes) are
written after the primary HDU. In this way the keywords of the the first
HDU can be used as metadata of the whole file (which may contain many
extensions, this is stipulated in the FITS standard). Usually the
primary HDU keywords contains the option names and values that the
program was run with. Because of this, Gnuastro's default HDU to read
data in a FITS file is the second (or '--hdu=1'). This error is commonly
caused when the FITS file wasn't created by Gnuastro or by a program
respecting this convention.
```