-
Notifications
You must be signed in to change notification settings - Fork 64
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.
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
Add back n_cpus argument in DeseqDataSet #226
Conversation
Thanks @vcabeli for this PR! Regarding the inference, IMO it's better not to require an |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, thanks!
pydeseq2/default_inference.py
Outdated
self._n_processes = utils.get_num_processes(n_cpus) | ||
self._backend = backend | ||
|
||
@property | ||
def n_cpus(self) -> int: # noqa: D102 | ||
return self._n_processes | ||
|
||
@n_cpus.setter | ||
def n_cpus(self, n_cpus: int) -> None: | ||
self._n_processes = n_cpus | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I might be missing something but wouldn't it be simpler to just set self.n_cpus = utils.get_num_processes(n_cpus)
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Right, without the abstract property it is indeed simpler to change the default inference's _n_processes
to n_cpus
pydeseq2/dds.py
Outdated
@@ -270,8 +278,20 @@ def __init__( | |||
self.logmeans = None | |||
self.filtered_genes = None | |||
|
|||
if n_cpus: | |||
n_cpus = get_num_processes(n_cpus) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't think we need to call get_num_processes
here, because it's already called when the DefaultInference
object is initialized. We could expect other inference implementations to handle the n_cpus
argument themselves.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We could expect other inference implementations to handle the n_cpus argument themselves
I think it's reasonable to expect a check during the __init__
, but should we expect another check when setting the parameter ? When inference
is specified, we would be overriding its n_cpus
param without checking it first.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We could add this check in the setter if we decide to keep it
7b85cb6
to
775b678
Compare
Reference Issue or PRs
Fixes #214
What does your PR implement? Be specific.
Adding back n_cpus as an optional keyword argument in DeseqDataSet, and adding
n_cpus
as an abstract property in the Inference class, in order to override the cpus of theinference
instance passed to DeseqDataSet (let me know if you prefer to not have that and just have a warning instead).