Feature/exe 1989 iterative leiden #188
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ptajvar
force-pushed
the
feature/exe-1989-iterative-leiden
branch
from
0602bb2 to
1e44d02
Compare
johandahlberg
approved these changes
Sep 17, 2024
| communities_graph = nx.from_edgelist(connected_communities) | ||
| for cc in nx.connected_components(communities_graph): | ||
| community_serie[community_serie.isin(cc)] = min(cc) | ||
| return edgelist, community_serie | ||
|
|
||
|
|
||
| def recover_technical_multiplets( |
There was a problem hiding this comment.
I think this function could benefit from being broken up into some smaller private functions to help a bit with readability.
ptajvar
force-pushed
the
feature/exe-1989-iterative-leiden
branch
from
61caad4 to
744aa90
Compare
4 tasks
ambarrio
approved these changes
Sep 24, 2024
Co-authored-by: Alvaro Martinez Barrio <[email protected]>
Co-authored-by: Alvaro Martinez Barrio <[email protected]>
Co-authored-by: Alvaro Martinez Barrio <[email protected]>
ludvigla
approved these changes
Oct 1, 2024
| @@ -32,7 +38,7 @@ def connect_components( | |||
| sample_name: str, | |||
| metrics_file: str, | |||
| multiplet_recovery: bool, | |||
| leiden_iterations: int = 10, | |||
| max_refinement_recursion_depth: int = 5, | |||
There was a problem hiding this comment.
I think 5 or 10 shouldn't make a huge difference. Perhaps it's a good idea to stick with this lower threshold of 5 for now and then we can evaluate whether this thresholds needs to be refined later.
…gelist-metrics-light removing components_modularity from edgelist_metrics
Co-authored-by: Alvaro Martinez Barrio <[email protected]>
… as an input to the graph step cli.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
To improve identification of potential technical multiplets, we are now implementing an iterative Leiden step where instead of a single leiden run on the entire sample, we run leiden iteratively. During each pass, if the subgraph is broken down into multiple communities the new communtities will be added to a queue to go through community detection again. This process goes on until a the subgraphs are no longer broken down into communities or a maximum depth given by "leiden_iterations" is reached.
In addition, the component names will now be generated using a hash of their pixels' barcodes and the generated components are deterministic. "components_modularity" is also now removed from edgelist_metrics as it is computationally expensive and the information is not frequently used.
Fixes: EXE-1989, EXE-1884, EXE-1979, EXE-2015
Type of change
Please delete options that are not relevant.
How Has This Been Tested?
The existing unit tests pass. The new graph step is run on multiple samples confirming intended detection of potential multiplets.
PR checklist: