Allow matrix multiplication in structured arrays

I'm using structured arrays to store atoms data produced by LAMMPS (I'm using a structured array that follows its format). I need to rotate the positions:

```
import numpy as np

transform = np.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]], dtype=np.float64)
dtype = np.dtype([("x", np.float64), ("y", np.float64), ("z", np.float64)])  # real case with more fields, integers, bools, strings

atoms = np.array(
    [
        (0.0, 0.0, 0.0),
        (1.0, 0.0, 0.0),
        (0.0, 1.0, 0.0),
        (1.0, 1.0, 1.0),
    ],
    dtype=dtype,
)

atoms[["x", "y", "z"]] = atoms[["x", "y", "z"]] @ transform.T
```

But this produces:

```
Traceback (most recent call last):
  File "c:\Users\acgc99\Desktop\rotation.py", line 16, in <module>
    atoms[["x", "y", "z"]] = atoms[["x", "y", "z"]] @ transform.T
                             ~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~
numpy._core._exceptions._UFuncNoLoopError: ufunc 'matmul' did not contain a loop with signature matching types (dtype([('x', '<f8'), ('y', '<f8'), ('z', '<f8')]), dtype('float64')) -> None
```

I can convert to unstructured arrays, but I guess that doing that change multiple times is not efficient when working with tens of millions of atoms.
_______________________________________________
NumPy-Discussion mailing list -- [email protected]
To unsubscribe send an email to [email protected]
https://mail.python.org/mailman3//lists/numpy-discussion.python.org
Member address: [email protected]