numpy.lib.format.descr_to_dtype

lib.format.descr_to_dtype(descr)

Returns a dtype based off the given description.

This is essentially the reverse of dtype_to_descr(). It will remove the valueless padding fields created by, i.e. simple fields like dtype(‘float32’), and then convert the description to its corresponding dtype.

Parameters:

descrobject

The object retrieved by dtype.descr. Can be passed to numpy.dtype() in order to replicate the input dtype.

Returns:

dtypedtype

The dtype constructed by the description.