image indexing for Brain and Nifti objects

[jeremymanning]

We can make Brain and Nifti objects indexable by making them iteratable objects: https://stackoverflow.com/questions/5359679/python-class-accessible-by-iterator-and-index

It looks like this involves adding __iter__ and __getitem__ methods:

• __iter__ returns the next item in the sequence (i.e. start a counter at 0 when the objects is initialized, and each time __iter__ is called, return object[counter+1] if len(object) < counter, or None otherwise, and then increment counter
• __getitem__ returns the given item(s)

[andrewheusser]

do we want the iterable to return a sliced bo, or the bo.data at a particular timepoint?

[andrewheusser]

in other words, a brain object containing a timepoint of data, or just the data itself?

[lucywowen]

I think the brain object containing the timepoint, since we can always calll bo.get_data after.

[andrewheusser]

i agree with lucy.. @jeremymanning what do you think?

[andrewheusser]

index both the timepoint and the electrode

[andrewheusser]

this is done

bo[0] #returns first time sample
bo[:5] #return first 5 time samples
bo[:, 0] #return first electrode
bo[:5, :5] #returns first 5 timesamples/elecs