Open
Conversation
* build_noun_graph – safely expand edge tuples * graph_to_dataframes – store embeddings as list objects * prune_graph – guard empty graphs + dtype-match source/target microsoft#1983
Author
@microsoft-github-policy-service agree |
1 similar comment
Author
@microsoft-github-policy-service agree |
This file contains hidden or 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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Tiny corpora +
CacheType.memorycrashed withValueError: Columns must be same length as key(and follow-on dtype issues).This PR hardens three spots:
build_noun_graph._extract_edges
edgescolumn intosource/targetviapd.DataFrame, padding bad rows and dropping NaNs leading to no broadcast error.graph_to_dataframes
listso the single-column assignment can’t broadcast-fail.prune_graph
astype(str)onsource/targetbefore merges to avoidfloat64 vs objectmismatch.Related Issues
Closes #1983
Proposed Changes
graphrag/index/operations/build_noun_graph/build_noun_graph.pygraphrag/index/operations/graph_to_dataframes.pygraphrag/index/operations/prune_graph.pytests/unit/indexing/graph/test_small_corpus_bug.pyChecklist
input/tiny.txt) withIndexingMethod.Fast&CacheType.memory.Additional Notes
These fixes touch only the small-corpus edge-cases - normal pipelines are unaffected.
Heads-up: the prompt files have been altered, can be discarded. Shout if you’d
like them dropped or squashed.