onnx_chainer.testing.input_generator.nonzero_increasing

onnx_chainer.testing.input_generator.nonzero_increasing(*shape, dtype=<class 'numpy.float32'>, bias=1e-07)

Returns a monotonically increasing ndarray for test inputs.

Similar to increasing but contains no zeros. Expected to be used for divisors.

Example:

>>> onnx_chainer.testing.input_generator.nonzero_increasing(3, 4)
array([[-3.0000000e+00, -2.5000000e+00, -1.9999999e+00, -1.4999999e+00],
       [-9.9999988e-01, -4.9999991e-01,  1.0000000e-07,  5.0000012e-01],
       [ 1.0000001e+00,  1.5000001e+00,  2.0000000e+00,  2.5000000e+00]],
      dtype=float32)
Parameters
  • shape (tuple of int) – The shape of the output array.

  • dtype (numpy.dtype) – The dtype of the output array.

  • bias (float) – The bias to avoid zero.

Returns

numpy.ndarray