mars.tensor.
nanargmin
Return the indices of the minimum values in the specified axis ignoring NaNs. For all-NaN slices ValueError is raised. Warning: the results cannot be trusted if a slice contains only NaNs and Infs.
ValueError
a (array_like) – Input data.
axis (int, optional) – Axis along which to operate. By default flattened input is used.
combine_size (int, optional) – The number of chunks to combine.
index_array – A tensor of indices or a single index value.
Tensor
See also
argmin, nanargmax
argmin
nanargmax
Examples
>>> import mars.tensor as mt
>>> a = mt.array([[mt.nan, 4], [2, 3]]) >>> mt.argmin(a).execute() 0 >>> mt.nanargmin(a).execute() 2 >>> mt.nanargmin(a, axis=0).execute() array([1, 1]) >>> mt.nanargmin(a, axis=1).execute() array([1, 0])