7. Table Data¶
Try it yourself
Download the data package (Try it yourself) if you wish to follow along and run the examples. Then
$ cd <path>/ad_usermanual/playground $ python
Then import core astrodata and the Gemini astrodata configurations.
>>> import astrodata >>> import gemini_instruments
7.1. Tables and Astrodata¶
Tables are stored as
Table class. FITS tables too
are represented in Astrodata as
Table and FITS headers are stored in
the NDAstroData .meta attribute. Most table access should be done
Table interface. The best reference on
Table is the
Astropy documentation itself. In this chapter we covers some common
examples to get the reader started.
astropy.table documentation can be found at: http://docs.astropy.org/en/stable/table/index.html
7.2. Operate on a Table¶
Let us open a file with tables. Some tables are associated with specific extensions, and there is one table that is global to the AstroData object.
>>> ad = astrodata.open('../playdata/N20170609S0154_varAdded.fits') >>> ad.info()
To access the global table named
To access the
OBJCAT table in the first extension
7.2.1. Column and Row Operations¶
Columns are named. Those names are used to access the data as columns. Rows are not names and are simply represented as a sequential list.
220.127.116.11. Read columns and rows¶
To get the names of the columns present in the table:
>>> ad.REFCAT.colnames ['Id', 'Cat_Id', 'RAJ2000', 'DEJ2000', 'umag', 'umag_err', 'gmag', 'gmag_err', 'rmag', 'rmag_err', 'imag', 'imag_err', 'zmag', 'zmag_err', 'filtermag', 'filtermag_err']
Then it is easy to request the values for specific columns:
>>> ad.REFCAT['zmag'] >>> ad.REFCAT['zmag', 'zmag_err']
To get the content of a specific row, row 10 in this case:
To get the content of a specific row(s) from a specific column(s):
>>> ad.REFCAT['zmag'] >>> ad.REFCAT['zmag'][4:10] >>> ad.REFCAT['zmag', 'zmag_err'][4:10]
18.104.22.168. Change values¶
Assigning new values works in a similar way. When working on multiple elements it is important to feed a list that matches in size with the number of elements to replace.
>>> ad.REFCAT['imag'] = 20.999 >>> ad.REFCAT['imag'][4:10] = [5, 6, 7, 8, 9, 10] >>> overwrite_col =  * len(ad.REFCAT) # a list of zeros, size = nb of rows >>> ad.REFCAT['imag_err'] = overwrite_col
22.214.171.124. Add a row¶
To append a row, there is the
add_row() method. The length of the row
should match the number of columns:
>>> new_row =  * len(ad.REFCAT.colnames) >>> ad.REFCAT.add_row(new_row)
126.96.36.199. Add a column¶
Adding a new column can be more involved. If you need full control, please see the AstroPy Table documentation. For a quick addition, which might be sufficient for your use case, we simply use the “dictionary” technique. Please note that when adding a column, it is important to ensure that all the elements are of the same type. Also, if you are planning to use that table in IRAF/PyRAF, we recommend not using 64-bit types.
>>> import numpy as np >>> new_column =  * len(ad.REFCAT) >>> # Ensure that the type is int32, otherwise it will default to int64 >>> # which IRAF 32-bit does not like. >>> new_column = np.array(new_column).astype(np.int32) >>> ad.REFCAT['my_column'] = new_column
If you are going to write that table back to disk as a FITS Bintable, then some additional headers need to be set. Astrodata will take care of that under the hood when the write method is invoked.
7.2.2. Selection and Rejection Operations¶
Normally, one does not know exactly where the information needed is located in a table. Rather some sort of selection needs to be done. This can also be combined with various calculations. We show two such examples here.
188.8.131.52. Select a table element from criterion¶
>>> # Get the magnitude of a star selected by ID number >>> ad.REFCAT['zmag'][ad.REFCAT['Cat_Id'] == '1237662500002005475'] >>> # Get the ID and magnitude of all the stars brighter than zmag 18. >>> ad.REFCAT['Cat_Id', 'zmag'][ad.REFCAT['zmag'] < 18.]
184.108.40.206. Rejection and selection before statistics¶
>>> t = ad.REFCAT # to save typing >>> # The table has "NaN" values. ("Not a number") We need to ignore them. >>> t['zmag'].mean() nan >>> # applying rejection of NaN values: >>> t['zmag'][np.where(~np.isnan(t['zmag']))].mean() 20.377306
7.2.3. Accessing FITS table headers directly¶
If for some reason you need to access the FITS table headers directly, here is how to do it. It is very unlikely that you will need this.
To see the FITS headers:
>>> ad.REFCAT.meta['header'] >>> ad.OBJCAT.meta['header']
To retrieve a specific FITS table header:
>>> ad.REFCAT.meta['header']['TTYPE3'] 'RAJ2000' >>> ad.OBJCAT.meta['header']['TTYPE3'] 'Y_IMAGE'
To retrieve all the keyword names matching a selection:
>>> keynames = [key for key in ad.REFCAT.meta['header'] if key.startswith('TTYPE')]
7.3. Create a Table¶
To create a table that can be added to an
AstroData object and eventually
written to disk as a FITS file, the first step is to create an Astropy
Let us first add our data to NumPy arrays, one array per column:
>>> import numpy as np >>> snr_id = np.array(['S001', 'S002', 'S003']) >>> feii = np.array([780., 78., 179.]) >>> pabeta = np.array([740., 307., 220.]) >>> ratio = pabeta / feii
Then build the table from that data:
>>> from astropy.table import Table >>> my_astropy_table = Table([snr_id, feii, pabeta, ratio], ... names=('SNR_ID', 'FeII', 'PaBeta', 'ratio'))
Now we append this Astropy
Table to a new
>>> # Since we are going to write a FITS, we build the AstroData object >>> # from FITS objects. >>> from astropy.io import fits >>> phu = fits.PrimaryHDU() >>> ad = astrodata.create(phu) >>> ad.append(my_astropy_table, name='MYTABLE') >>> ad.info() >>> ad.MYTABLE >>> ad.write('new_table.fits')