Repository files navigation

Trait types for NumPy, SciPy and friends

Provide a reference implementation of trait types for common data structuresused in the scipy stack such as

• numpy arrays
• pandas andxarray data structures

which are out of the scope of the maintraitletsproject but are a common requirement to build applications with traitlets incombination with the scipy stack.

Another goal is to create adequate serialization and deserialization routinesfor these trait types to be used with theipywidgetsproject (to_json andfrom_json). These could also return a list of binarybuffers as allowed by the current messaging protocol.

Usingpip:

Make sure you havepip installed and run:

pip install traittypes

Usingconda:

conda install -c conda-forge traittypes

traittypes extends thetraitlets library with an implementation of trait types for numpy arrays, pandas dataframes, pandas series, xarray datasets and xarray dataarrays.

• traittypes works around some limitations with numpy array comparison to only trigger change events when necessary.
• traittypes also extends the traitlets API for adding custom validators to constrain proposed values for the attribute.

For a general introduction totraitlets, check out thetraitlets documentation.

fromtraitletsimportHasTraits,TraitErrorfromtraittypesimportArraydefshape(*dimensions):defvalidator(trait,value):ifvalue.shape!=dimensions:raiseTraitError('Expected an of shape %s and got and array with shape %s'% (dimensions,value.shape))else:returnvaluereturnvalidatorclassFoo(HasTraits):bar=Array(np.identity(2)).valid(shape(2,2))foo=Foo()foo.bar= [1,2]# Should raise a TraitError