Traitlets types for NumPy, SciPy and friends
Clone or download
SylvainCorlay Merge pull request #29 from vidartf/fix-sentinel
Copy Sentinel implementation from traitlets
Latest commit 609b402 Jul 18, 2018

Scipy Trait Types

Build Status Documentation Status

Trait types for NumPy, SciPy and friends


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

which are out of the scope of the main traitlets project but are a common requirement to build applications with traitlets in combination with the scipy stack.

Another goal is to create adequate serialization and deserialization routines for these trait types to be used with the ipywidgets project (to_json and from_json). These could also return a list of binary buffers as allowed by the current messaging protocol.


Using pip:

Make sure you have pip installed and run:

pip install traittypes

Using conda:

conda install -c conda-forge traittypes


traittypes extends the traitlets 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 constained proposed values for the attribute.

For a general introduction to traitlets, check out the traitlets documentation.

Example usage with a custom validator

from traitlets import HasTraits, TraitError
from traittypes import Array

def shape(*dimensions):
    def validator(trait, value):
        if value.shape != dimensions:
            raise TraitError('Expected an of shape %s and got and array with shape %s' % (dimensions, value.shape))
            return value
    return validator

class Foo(HasTraits):
    bar = Array(np.identity(2)).valid(shape(2, 2))
foo = Foo() = [1, 2]  # Should raise a TraitError