Skip to content

update var by dist normalization #1065

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
LLehner wants to merge 2 commits into
base: main
Choose a base branch
from
+3 −10
Open

update var by dist normalization #1065

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
80948fa
update var by dist normalization
LLehner Nov 20, 2025
f7b8b2e
fix pre-commit
LLehner Nov 20, 2025
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
13 changes: 3 additions & 10 deletions src/squidpy/tl/_var_by_distance.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,6 @@
from scanpy import logging as logg
from sklearn.metrics import DistanceMetric
from sklearn.neighbors import KDTree
from sklearn.preprocessing import MinMaxScaler

from squidpy._docs import d
from squidpy._utils import NDArrayA
Expand Down Expand Up @@ -92,7 +91,6 @@ def var_by_distance(
raise TypeError(f"Invalid type for library_key: {type(library_key)}.")

batch_design_matrices = {}
max_distances = {}
anchor_col_id = 2
# iterate over slide + anchor combinations (anchor only possible as well)
combinations: Iterator[tuple[str, str | None]] | zip[tuple[list[int | float], str]] # mypy
Expand Down Expand Up @@ -146,12 +144,11 @@ def var_by_distance(
else:
df["obs"] = adata.obs_names

# store dataframes by (slide, anchor) combination and also the corresponding maximum distance for normalization
# store dataframes by (slide, anchor) combinationq
Copy link
Member

@timtreis timtreis Nov 22, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

typo

batch_design_matrices[(batch_var, anchor_var)] = df
max_distances[(batch_var, anchor_var)] = df[anchor_var].max()

# normalize euclidean distances by slide
batch_design_matrices_ = _normalize_distances(batch_design_matrices, anchor, batch, max_distances)
batch_design_matrices_ = _normalize_distances(batch_design_matrices, anchor)

# combine individual data frames
# merge if multiple anchor points are used but there is no separation by slides
Expand Down Expand Up @@ -244,8 +241,6 @@ def _get_coordinates(
def _normalize_distances(
mapping_design_matrix: dict[tuple[Any | None, Any], pd.DataFrame],
anchor: str | list[str],
slides: list[str] | list[None],
mapping_max_distances: dict[tuple[Any | None, Any], float],
) -> list[pd.DataFrame]:
"""Normalize distances to anchor."""
if not isinstance(anchor, list):
Expand All @@ -259,7 +254,5 @@ def _normalize_distances(

# normalize each slide separately will result in all conditions having a maximum distance of 1, which makes them comparable across all distances
for (_, anchor_point), design_matrix in mapping_design_matrix.items():
scaler = MinMaxScaler()
scaler.fit(design_matrix[[anchor_point]].values)
design_matrix[anchor_point] = scaler.transform(design_matrix[[anchor_point]].values)
design_matrix[anchor_point] = design_matrix[anchor_point] / design_matrix[anchor_point].max()
return list(mapping_design_matrix.values())