Added trials parameter to leiden #790
Conversation
…ogic-native 1.0.0 or higher to use it
|
Deploy Preview for graspologic failed. Built with commit 0f6a1a9 https://app.netlify.com/sites/graspologic/deploys/60a43eb4aeb92300078cea94 |
| if not isinstance(trials, int): | ||
| raise TypeError("trials must be a positive integer") | ||
| if trials < 1: | ||
| raise ValueError("trials must be a positive integer") |
There was a problem hiding this comment.
I know we've talked about input checking at various times - I've always thought it could be helpful to use things like https://scikit-learn.org/stable/modules/generated/sklearn.utils.check_scalar.html going forward, just to be more concise and consistent. I don't care for the purposes of this PR but thought I'd bring it up.
| @@ -277,6 +278,13 @@ def leiden( | |||
| if it is found to be a directed graph. If you know it is undirected and wish to | |||
| avoid this scan, you can set this value to ``False`` and only the lower triangle | |||
| of the adjacency matrix will be used to generate the weighted edge list. | |||
| trials : int | |||
There was a problem hiding this comment.
I also don't want to let this hold up the PR, but this seems like something we should add to the stack of #755 and consider how we name parameters like this in the future. This exact thing happens in match and cluster at a minimum and there we call it n_init.
|
for some reason netlify build failed with |
Added trials parameter to leiden and set a new requirement of graspologic-native 1.0.0 or higher to use it.