Chris@87: """
Chris@87: =============
Chris@87: Masked Arrays
Chris@87: =============
Chris@87: 
Chris@87: Arrays sometimes contain invalid or missing data.  When doing operations
Chris@87: on such arrays, we wish to suppress invalid values, which is the purpose masked
Chris@87: arrays fulfill (an example of typical use is given below).
Chris@87: 
Chris@87: For example, examine the following array:
Chris@87: 
Chris@87: >>> x = np.array([2, 1, 3, np.nan, 5, 2, 3, np.nan])
Chris@87: 
Chris@87: When we try to calculate the mean of the data, the result is undetermined:
Chris@87: 
Chris@87: >>> np.mean(x)
Chris@87: nan
Chris@87: 
Chris@87: The mean is calculated using roughly ``np.sum(x)/len(x)``, but since
Chris@87: any number added to ``NaN`` [1]_ produces ``NaN``, this doesn't work.  Enter
Chris@87: masked arrays:
Chris@87: 
Chris@87: >>> m = np.ma.masked_array(x, np.isnan(x))
Chris@87: >>> m
Chris@87: masked_array(data = [2.0 1.0 3.0 -- 5.0 2.0 3.0 --],
Chris@87:       mask = [False False False  True False False False  True],
Chris@87:       fill_value=1e+20)
Chris@87: 
Chris@87: Here, we construct a masked array that suppress all ``NaN`` values.  We
Chris@87: may now proceed to calculate the mean of the other values:
Chris@87: 
Chris@87: >>> np.mean(m)
Chris@87: 2.6666666666666665
Chris@87: 
Chris@87: .. [1] Not-a-Number, a floating point value that is the result of an
Chris@87:        invalid operation.
Chris@87: 
Chris@87: """
Chris@87: from __future__ import division, absolute_import, print_function
Chris@87: 
Chris@87: __author__ = "Pierre GF Gerard-Marchant ($Author: jarrod.millman $)"
Chris@87: __version__ = '1.0'
Chris@87: __revision__ = "$Revision: 3473 $"
Chris@87: __date__     = '$Date: 2007-10-29 17:18:13 +0200 (Mon, 29 Oct 2007) $'
Chris@87: 
Chris@87: from . import core
Chris@87: from .core import *
Chris@87: 
Chris@87: from . import extras
Chris@87: from .extras import *
Chris@87: 
Chris@87: __all__ = ['core', 'extras']
Chris@87: __all__ += core.__all__
Chris@87: __all__ += extras.__all__
Chris@87: 
Chris@87: from numpy.testing import Tester
Chris@87: test = Tester().test
Chris@87: bench = Tester().bench