Usage of python in CASA

In this chapter we review some issues with programming in python in the CASA environment

CASA vs numpy

Python indices are 0 based, like in C/C++. CASA arays are also indexed 0-based, but for those familiar with python’s numpy and masking arrays, this is where all similarity ends. Image data in CASA are columnn major, as in Fortran, where in numpy there are row major, as in C. Masking arrays in casa and numpy’s masking have a reversed logic (mask=True means a bad data point in numpy, but a good one in CASA)

Our standard example below is an array with 2 planes, 3 rows and 4 columns. The values in the array will be i + 10*j + 100*k, where i counts the columns, j the rows, and k the planes, all 0-based indexed. This the first value in the array is 0 (000), the last one 123. You can also find a file cube432.fits in the ADMIT data distribution.

import numpy    as np
import as ma

a = np.arange(24).reshape(2,3,4)
print a.shape, a[1,2,3], a[1][2][3]              # should print: (2, 3, 4) 23 23

# re-assign values based on their (i,j,k) index
for k in range(a.shape[0]):
    for j in range(a.shape[1]):
        for j in range(a.shape[2]):
            a[k,j,i] = i+10*j+100*k

# mask every third number bad
b = ma.masked_where(a%3==0,a)
print a.size, b.count()                          # should print:  24 16

We will now see how masking in python and CASA is different.

CODING style

When reading the help files within CASA, there are a few bad habits you can pick up, which may seem convenient, but should not be used if you want your code to be more portable outside of CASA:

In random order

  • True and False are the python literals for the two boolean values, but you will see both true and t being used in CASA examples. They are defined for convenience within CASA, but if you ever want to use your code outside of CASA, this will obviously cause problems. There is no reason to not use the official names (if not just for your colorizing editor to recognize them and color them appropriately), so use the original python literals.
  • more to come


For an ADMIT developer environment (the case where your shell has an $ADMIT environment variable and you would have loaded this by sourcing the appropriate admit_start.[c]sh script), your CASA environment has also been modified to include not only $CASA_ROOT/bin in your $PATH, but also $CASA_ROOT/lib/casa/bin.

If you want to build documentation, you will also have had to install pip and sphinx. The make pip target in the $ADMIT directory should do this for most installations.