Utility for sharing numpy arrays between processes. This can be useful if you have a background process performing a calculation and main process on an API which can query the array.
Usage:
from np_shared_array import create_shared_array, SharedNumpyArray
from multiprocessing import Process
def worker(shared_array: SharedNumpyArray):
array = shared_array.to_numpy()
with shared_array.lock:
array[0][0] = 1
def main():
shape = [10, 10]
shared_array = create_shared_array(shape)
array = shared_array.to_numpy()
p = Process(
target=worker,
args=(shared_array,)
)
p.start()
p.join()
with shared_array.lock:
print(array[0][0])
if __name__ == "__main__":
main()