Assertion and Testing

Chainer provides some facilities to make debugging easy.

Type checking utilities

Function uses a systematic type checking of the chainer.utils.type_check module. It enables users to easily find bugs of forward and backward implementations. You can find examples of type checking in some function implementations.

chainer.utils.type_check.Expr Abstract syntax tree of an expression.
chainer.utils.type_check.expect Evaluates and tests all given expressions.
chainer.utils.type_check.TypeInfo Type information of an input/gradient array.
chainer.utils.type_check.TypeInfoTuple Type information of input/gradient tuples.

Gradient checking utilities

Most function implementations are numerically tested by gradient checking. This method computes numerical gradients of forward routines and compares their results with the corresponding backward routines. It enables us to make the source of issues clear when we hit an error of gradient computations. The chainer.gradient_check module makes it easy to implement the gradient checking.

chainer.gradient_check.check_backward Test backward procedure of a given function.
chainer.gradient_check.numerical_grad Computes numerical gradient by finite differences.

Standard Assertions

The assertions have same names as NumPy’s ones. The difference from NumPy is that they can accept both numpy.ndarray and cupy.ndarray.

chainer.testing.assert_allclose Asserts if some corresponding element of x and y differs too much.

Function testing utilities

Chainer provides some utilities for testing its functions.

chainer.testing.unary_math_function_unittest Decorator for testing unary mathematical Chainer functions.