Mask numpy array with a range of values

Question

I have a numpy structured array with one of the columns called 'time' I want to restrict my data where the 'time' column is in a certain range. I tried this:

time_restricted_data = Data[ (Data['time'] > 0.6) & (Data['time'] < 0.7) ]

But that returned all the data where Data['time'] > 0.6.

Any suggestions?

I'm a little confused what Data is...you're indexing it with a string and then later your indexing it with something which (probably) isn't a string. Could provide a little more context code in your example? — mgilson
– mgilson, Commented May 1, 2012 at 14:55
@mgilson numpy arrays can be indexed by arrays of bools, the effect of which is to extract a subset of the array corresponding to the 'true' bool elements. I'm not familiar with structured arrays (which can be indexed by strings) but I have had problems indexing a normal array A using a bool array B, in the case where B needs one or more dimensions expanded to fit the shape of A. — greggo
– greggo, Commented May 1, 2012 at 15:03
@greggo -- I understand that numpy arrays can be indexed by arrays of bools. I just think that we need to know how he has it set up so that Data can be indexed by a string ("time") and (presumably) an array of bools at the same time...It's possible that he just needs to do time_restricted_data = Data["time"][...] (e.g. if Data is really a dict holding numpy arrays and he forgot to mention that...) — mgilson
– mgilson, Commented May 1, 2012 at 15:07

segasai · Accepted Answer · 2012-05-01 15:29:43Z

1

Something is wrong in your example, i.e. I can't confirm it, and I think what you wrote should work as expected:

In [19]: Data = np.zeros(100,dtype=[('time', np.float), ('y',np.float)])

In [20]: Data['y'] = np.random.uniform(size=100)

In [21]: Data['time'] = np.random.uniform(size=100)

In [22]: print Data[ (Data['time'] > 0.6) & (Data['time'] < 0.7) ]
[(0.6309334093696576, 0.5898588768194092)
 (0.6026040512366535, 0.4260650141076221)
 (0.6587399844526572, 0.033397798015253444)
 (0.6863639946779522, 0.67002523603246)
 (0.6522035987367735, 0.948019085443445)
 (0.6809894254849801, 0.5131390279565994)
 (0.6311277013562147, 0.5746610745753917)
 (0.6324174554481182, 0.8587836614681397)
 (0.6542221804687635, 0.9706926940115863)
 (0.671321726341415, 0.7446681474117195)]

answered May 1, 2012 at 15:29

segasai

8,5581 gold badge31 silver badges40 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

mgilson Over a year ago

Thanks, this explains how you can index a np array by a string and boolean values -- I'd never seen this done before, although it seems like it could be quite useful.

jlconlin Over a year ago

Well I tried again what I had tried before and it works. Your solution is correct. I must have had a typo somewhere before.

Collectives™ on Stack Overflow

Mask numpy array with a range of values

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related