Some confusion working in absolute coordinates in ACAImage

[jeanconn]

Reviewing the docs:

>>> from chandra_aca.aca_image import ACAImage
>>> im4 = np.arange(16).reshape(4, 4)
>>> a = ACAImage(im4, row0=10, col0=20) 
>>> a
<ACAImage row0=10 col0=20
array([[ 0,  1,  2,  3],
       [ 4,  5,  6,  7],
       [ 8,  9, 10, 11],
       [12, 13, 14, 15]])>
>>> a[3, 3]
15
>>> a.aca[13, 23]
15

That's fine. Here's where I get a little confused:

In [90]: a.aca[13, 16]
Out[90]: 12
In [91]: a.aca[13, 20]
Out[91]: 12
In [92]: a.aca[9, 16]
Out[92]: 12

It seems like these mappings back into the original array aren't quite absolute? I expect an error somewhere here.

[jeanconn]

Ah, I think I get it... the aca_coords are subtracting off row0, col0, so my examples are just benefiting from negative indexing back in to the original array. It might be good to check the bounds on the aca coordinates of the ACAImage in question and throw an error instead of allowing the negative indexing.

[taldcroft]

Agreed an exception for out-of-bounds absolute indexing makes sense. Not sure if practical issues will come up (performance or other cases like slicing that goes out of bounds).