Cannot cast array(int64, 1d, C) to int32

Numba Support: How do I do ...?
1

Hello!

I am new to Numba and I have been having some troubles with the data types, I tried to convert an np.int64 to np.int32, I kept tracing and discovered that the problem was the conversion from int64, however I have no clue of what could I try to solve it:
Cannot cast array(int64, 1d, C) to int32: %".7980" = load {i8*, i8*, i64, i64, i64*, [1 x i64], [1 x i64]}, {i8*, i8*, i64, i64, i64*, [1 x i64], [1 x i64]}* %"$350binary_subscr.172"

I have made an example:
import numpy as np

import numba

from numba import jit

@jit(nopython=True,nogil=True)   

def t_corrido_i():

    grafo_e=np.array([[0,1,12],[0,2,12],[0,3,12]],dtype=np.int32)

    tarefas_pj=np.array([ 0,  9, 13, 15, 16])

    requere_nodes=grafo_e[np.where(grafo_e[:,1]==3)][:,0]

    # Alternative for np.isin for numba

    requere_no_outro_proc=np.array([x in set(requere_nodes) for x in tarefas_pj])

    ix_nos_requeridos=np.int32(np.where(requere_no_outro_proc)[0])

t_corrido_i()

Does anyone have a clue of what could that be?

2

hi there!
Numba thinks you are trying to convert the int64 array to an int32 scalar, that’s why it complains.

you can achieve the conversion to int32 by writing ix_nos_requeridos=np.where(requere_no_outro_proc)[0].astype(np.int32)

cheers,
Luk

3

Thanks so much! It worked out perfectly!

4

Opened numpy constructors aren't typing correctly, invalid cast issued · Issue #6656 · numba/numba · GitHub to track, it should have been caught earlier.