Selecting a subset of an astropy table

Question

I am trying to create a subset of an astropy table.

I am told I need to use numpy but when I access the table via numpy I don't seem to get a 2d numpy table and am confused.

The table I am trying to create a subset comes from the GAMA Survey

I have tried

import numpy as np
from astropy.table import Table
t = Table.read('GAMA_Data/InputCatA.fits')
ta = np.array(t)

I wish to perform a number of subsets like

t['PETROMAG_I'] > 17.5)

I received some advice that I need to use numpy. I can determine that 'PETROMG_I' is column 16 with

print(t.colnames.index('PETROMAG_I'))

But do not see how to do a subset on either the table or numpy array

 print('Source Table : '+str(len(ta)))
 print('Shape : '+str(ta.shape))
 print(ta)

Gives

Source Table : 960510
Shape : (960510,)
[(     0, 588848900971299297, 145.46972201, 0.65141891,           68987912448,        0, 3.0191224, 20.57372 , 20.748629, 4503874773745664, 0.3678997 , 2.9388053, 1.2919843 , 28.094984, 21.347776, 19.99194 , 19.40016 , 18.74607 , 22.722446, 21.583874, 19.939884, 19.270464, 18.782948,  9843, 756)
 (     1, 588848900971299302, 145.47892146, 0.63515137,       68987912448,        0, 3.3755734, 21.073107, 21.055092, 4503874773745664, 0.37012857, 3.619758 , 1.5749472 , 21.076248, 20.432001, 20.08754 , 19.929377, 20.802927, 21.622843, 20.553198, 20.045462, 19.874853, 19.653866,  9843, 756)

Answer 1

Use:

mask = t['PETROMAG_I'] > 17.5  # Makes a boolean selection mask (numpy array)
t_new = t[mask]  # Makes a new astropy Table

This will give you an astropy Table with the selected rows. This is documented in an example near the bottom of the section here: https://docs.astropy.org/en/stable/table/access_table.html#accessing-data .

Answer 2

From your printout it look like it was a list of tuples , but probably it is actually a structured Numpy array and 960510 is the number of records (rows) in this table.

Unfortunately, neither field names nor their types can be seen here, but try to execute print(ta.dtypes) .

Most likely you will see somethig like:

dtype([('Name_1', '<U10'), ('Name_2', '<f8'), ...])

a list of pairs (field name, field type), for details see Numpy documentation.

Then, to access one particular column (in your case PETROMAG_I ), run:

print(ta['PETROMAG_I'])

And to generate your subset, try:

mySubset = ta[ta['PETROMAG_I'] > 17.5]

Another option : Use to_pandas() method of astropy , to convert your Table into a Pandas DataFrame. DataFrames are much more "user friendly" than Numpy arays.