fix: gracefully handle oversized window_size with warning and adjustment

Replace hard error when window_size exceeds available SNPs with a warning,
and automatically adjust window_size to the maximum valid value so that
computation can proceed.

Signed-off-by: Suhrid Marwah <suhridmarwah07@gmail.com>

Author: suhr25

malariagen_data/anoph/fst.py

@@ -1,3 +1,4 @@
+import warnings
 from typing import Tuple, Optional
 
 import numpy as np
@@ -81,20 +82,23 @@ def _fst_gwss(
                 chunks=chunks,
             ).compute()
 
+        n_snps = len(pos)
+        if window_size > n_snps:
+            warnings.warn(
+                f"window_size ({window_size}) is larger than the number of SNP sites "
+                f"available ({n_snps}); adjusting window_size to {n_snps}.",
+                UserWarning,
+                stacklevel=2,
+            )
+            window_size = n_snps
+
         with self._spinner(desc="Compute Fst"):
             with np.errstate(divide="ignore", invalid="ignore"):
                 fst = allel.moving_hudson_fst(ac1, ac2, size=window_size)
                 # Sometimes Fst can be very slightly below zero, clip for simplicity.
                 fst = np.clip(fst, a_min=clip_min, a_max=1)
                 x = allel.moving_statistic(pos, statistic=np.mean, size=window_size)
 
-        if len(x) == 0:
-            raise ValueError(
-                f"No Fst windows could be computed: window_size={window_size!r} is "
-                f"larger than the number of SNP sites available ({len(pos)}) in the "
-                "selected region. Try reducing window_size or selecting a larger region."
-            )
-
         results = dict(x=x, fst=fst)
 
         return results
@@ -103,7 +107,9 @@ def _fst_gwss(
     @doc(
         summary="""
             Run a Fst genome-wide scan to investigate genetic differentiation
-            between two cohorts.
+            between two cohorts. If window_size exceeds the number of available
+            SNP sites, a UserWarning is issued and window_size is automatically
+            reduced to the number of available sites.
         """,
         returns=dict(
             x="An array containing the window centre point genomic positions",

tests/anoph/test_fst.py

@@ -142,15 +142,15 @@ def test_fst_gwss(fixture, api: AnophelesFstAnalysis):
 
 @parametrize_with_cases("fixture,api", cases=".")
 def test_fst_gwss_window_size_too_large(fixture, api: AnophelesFstAnalysis):
-    # Use a window_size larger than the number of available SNPs to trigger the
-    # ValueError guard added in _fst_gwss.
+    # When window_size exceeds available SNPs, a UserWarning must be issued and
+    # the function must still return a valid result using the adjusted window_size.
     all_sample_sets = api.sample_sets()["sample_set"].to_list()
     all_countries = api.sample_metadata()["country"].dropna().unique().tolist()
     countries = random.sample(all_countries, 2)
     cohort1_query = f"country == {countries[0]!r}"
     cohort2_query = f"country == {countries[1]!r}"
-    with pytest.raises(ValueError, match="window_size"):
-        api.fst_gwss(
+    with pytest.warns(UserWarning, match="window_size"):
+        x, fst = api.fst_gwss(
             contig=random.choice(api.contigs),
             sample_sets=all_sample_sets,
             cohort1_query=cohort1_query,
@@ -159,6 +159,10 @@ def test_fst_gwss_window_size_too_large(fixture, api: AnophelesFstAnalysis):
             window_size=10_000_000,  # far larger than any fixture SNP count
             min_cohort_size=1,
         )
+    assert isinstance(x, np.ndarray)
+    assert isinstance(fst, np.ndarray)
+    assert len(x) > 0
+    assert x.shape == fst.shape
 
 
 @parametrize_with_cases("fixture,api", cases=".")