I have an LDAC fits catalog which in a Python code I need to add the elements of two arrays as two new columns to it.
I open the original catalog in python:
from astropy.io import fits
from astropy.table import Table
import astromatic_wrapper as aw
cat1='catalog.cat'
hdulist1 =fits.open(cat1)
data1=hdulist1[1].data
The two arrays are ready and called ra and dec. I give them the key name, format and other needed info and invert them to columns. Finally, I join the two new columns to the original table (Checking newtab.columns and newtab.data shows that the new columns are attached successfully).
racol=fits.Column(name = 'ALPHA_J2000', format = '1D', unit = 'deg', disp = 'F11.7',array=ra)
deccol=fits.Column(name = 'DELTA_J2000', format = '1D', unit = 'deg', disp = 'F11.7',array=dec)
cols = fits.ColDefs([racol, deccol])
tbhdu = fits.BinTableHDU.from_columns(cols)
orig_cols= data1.columns
newtab = fits.BinTableHDU.from_columns(cols + orig_cols)
When I save the new table into a new catalog:
newtab.writeto('newcatalog.cat')
it is not in the format that I need. If I look into the description of each catalog with
ldacdes -i
I see for catalog.cat :
> Reading catalog(s)
------------------Catalog information----------------
Filename:..............catalog.cat
Number of segments:....3
****** Table #1
Extension type:.........(Primary HDU)
Extension name:.........
****** Table #2
Extension type:.........BINTABLE
Extension name:.........OBJECTS
Number of dimensions:...2
Number of elements:.....24960
Number of data fields...23
Body size:..............4442880 bytes
****** Table #3
Extension type:.........BINTABLE
Extension name:.........FIELDS
Number of dimensions:...2
Number of elements:.....1
Number of data fields...4
Body size:..............28 bytes
> All done
and for the new one:
> Reading catalog(s)
------------------Catalog information----------------
Filename:..............newcatalog.cat
Number of segments:....2
****** Table #1
Extension type:.........(Primary HDU)
Extension name:.........
****** Table #2
Extension type:.........BINTABLE
Extension name:.........
Number of dimensions:...2
Number of elements:.....24960
Number of data fields...25
Body size:..............4842240 bytes
> All done
As seen above, in the original catalog catalog.cat there are three tables and I tried to add two columns to the OBJECTS table.
I need that newcatalog.cat also keeps the same structure which is required by other programs, but it does not have the OBJECTS table and considering the "Number of elements" and the "Number of data fields" the newtab is saved into the Table #2.
Is there any solution for controlling the output fits catalog format?
Thank you for your help and I hope that I could structure my very first question on stackoverflow properly .
I don't know specifically about the LDAC format, but from your example file catalog.cat, it appears to be a multi-extension FITS file. That is, each table is stored in a separate HDU (as is typical for any file containing multiple tables with different sets of columns).
When you do something like
newtab = fits.BinTableHDU.from_columns(cols + orig_cols)
newtab.writeto('newcatalog.cat')
You're just creating a single new binary table HDU and writing that HDU to a file by itself (along with the mandatory primary HDU). What you really want is to take the same HDU structure as the original file and replace the existing table HDU with the one to which you added new columns.
Creating multi-extension FITS is discussed some here, but you don't even need to recreate the full HDU structure from scratch. The HDUList object return from fits.open is just a list of HDUs that can be manipulated like a normal Python list (with some extensions, for example, to support indexing by EXTNAME) and written out to a file:
hdulist = fits.open(cat1)
hdulist['OBJECTS'] = newtab
hdulist.writeto('newcatalog.cat')