Describe the bug
Running write_correlations for a large dataset for Python versions < 3.7 can result in multiprocessing errors during filtering. This appears to be fixed by this PR on cPython.
The error looks something like this:
Traceback (most recent call last):
File "python3.6/multiprocessing/queues.py", line 240, in _feed
send_bytes(obj)
File "python3.6/multiprocessing/connection.py", line 200, in send_bytes
self._send_bytes(m[offset:offset + size])
File "python3.6/multiprocessing/connection.py", line 393, in _send_bytes
header = struct.pack("!i", n)
struct.error: 'i' format requires -2147483648 <= number <= 2147483647
Additional context
I will leave this here as a note: it is worth updating to Python 3.7 or greater. However, I could/should have a serial option for max_workers=1 rather than using multiprocessing for 1 worker so that this can be worked-around without having to update Python.
Describe the bug Running
write_correlations
for a large dataset for Python versions < 3.7 can result in multiprocessing errors during filtering. This appears to be fixed by this PR on cPython.The error looks something like this:
Additional context I will leave this here as a note: it is worth updating to Python 3.7 or greater. However, I could/should have a serial option for
max_workers=1
rather than using multiprocessing for 1 worker so that this can be worked-around without having to update Python.