Skip to content

WIP: Let Dask pick automatic chunk size along train/time axis #322

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
takluyver wants to merge 1 commit into
base: master
Choose a base branch
from
Open

WIP: Let Dask pick automatic chunk size along train/time axis #322

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 4 additions & 11 deletions extra_data/keydata.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -373,17 +373,10 @@ def dask_array(self, labelled=False):
chunk_shape = (chunk_dim0,) + chunk.dataset.shape[1:]
itemsize = chunk.dataset.dtype.itemsize

# Find chunk size of maximum 2 GB. This is largely arbitrary:
# we want chunks small enough that each worker can have at least
# a couple in memory (Maxwell nodes have 256-768 GB in late 2019).
# But bigger chunks means less overhead.
# Empirically, making chunks 4 times bigger/smaller didn't seem to
# affect speed dramatically - but this could depend on many factors.
# TODO: optional user control of chunking
limit = 2 * 1024 ** 3
while np.product(chunk_shape) * itemsize > limit and chunk_dim0 > 1:
chunk_dim0 //= 2
chunk_shape = (chunk_dim0,) + chunk.dataset.shape[1:]
# Let Dask pick chunk size along the first (trains/time) dimension.
# It will aim for a chunk size of 128 MiB by default, though users can
# configure this with a setting called 'array.chunk-size'.
chunk_shape = ("auto",) + chunk.dataset.shape[1:]

chunks_darrs.append(
da.from_array(
Expand Down