shimao
shimao
I sent PR#10. Pls merge. Thanks
This way. ``` import numpy as np import multiprocessing from joblib import Parallel, delayed from skimage.feature import peak_local_max def _process_on_multi_core(x: np.ndarray) -> np.ndarray: n_core = multiprocessing.cpu_count() n_sub_block_y = x.shape[0]//n_core n_sub_block_x...
@datajoely Can't you reproduce it by creating a node that just uses this function? What information do I need to provide to get started?
@datajoely When you say work with `SequentialRunner`, do you mean that parallel operations will be done using joblib? I am simply running it with `kedro run`, so I am running...
@datajoely No. I call `_process_on_multi_core` from node function as sub function.
No errors occur. There is no stacktrace because the calculation is not performed on multiple cores, but only on one core.