CODEWITHSUNDEEP
pythonscipyintel-mkl
Reza
Reza

Reputation: 398

MKL: ValueError: Matrix data types must be in concordance; int32 and int32 provided

from scipy import sparse
from sparse_dot_mkl import dot_product_mkl

data =[1,1,1,1]
Row=[1, 100, 2, 200]
Col=[53, 21, 18, 32]
N=210

A=sparse.csc_matrix((data, (Row,Col)),  shape=(N,N) )
M=A.copy()
dot_product_mkl(A,A)

I get this error:

> ValueError                                Traceback (most recent call
> last) <ipython-input-10-fa6e7f71b940> in <module>
> ----> 1 dot_product_mkl(A,A)
> 
> ~\anaconda3\lib\site-packages\sparse_dot_mkl\sparse_dot.py in
> dot_product_mkl(matrix_a, matrix_b, cast, copy, reorder_output, dense,
> debug, out, out_scalar)
>      56 
>      57     elif num_sparse == 2:
> ---> 58         return _sds(matrix_a, matrix_b, cast=cast, reorder_output=reorder_output, dense=dense)
>      59 
>      60     # SPARSE (DOT) VECTOR #
> 
> ~\anaconda3\lib\site-packages\sparse_dot_mkl\_sparse_sparse.py in
> _sparse_dot_sparse(matrix_a, matrix_b, cast, reorder_output, dense)
>     127 
>     128     # Check dtypes
> --> 129     matrix_a, matrix_b = _type_check(matrix_a, matrix_b, cast=cast)
>     130 
>     131     t = debug_timer()
> 
> ~\anaconda3\lib\site-packages\sparse_dot_mkl\_mkl_interface\_common.py
> in _type_check(matrix_a, matrix_b, cast)
>     624         err_msg = "Matrix data types must be in concordance; {a} and {b} provided".format(a=matrix_a.dtype,
>     625                                                                                           b=matrix_b.dtype)
> --> 626         raise ValueError(err_msg)
>     627 
>     628 def _mkl_scalar(scalar, complex_type, double_precision):
> 
> ValueError: Matrix data types must be in concordance; int32 and int32
> provided

The two matrices are copies of each other so I don't understand what this message means.

Upvotes: 0

Views: 410

Answers (1)

Rahila T - Intel
Rahila T - Intel

Reputation: 862

This package only works with float or complex float data.

The below link will help you to understand sparse dot product in detail.

https://pypi.org/project/sparse-dot-mkl/

Here it is clearly mentioned that, this package only works with float or complex float data. cast=True will convert data to double-precision floats or complex floats by making an internal copy if necessary. If A and B are both single-precision floats or complex floats they will be used as is. cast=False will raise a ValueError if the input arrays are not both double-precision or both single-precision. This defaults to False on the principle that potentially unsafe dtype conversions should not occur without explicit instruction.

Upvotes: 1

Related Questions