Merge branch 'master' into fix/h12-multi-panel-param-forwarding

Author: jonbrenas

malariagen_data/anoph/base.py

@@ -562,6 +562,13 @@ def _sample_set_has_unrestricted_use(self, *, sample_set: str):
         release_manifest_df = self._read_sample_sets_manifest(
             single_release=sample_set_release
         )
+
+        if "unrestricted_use" not in release_manifest_df.columns:
+            raise ValueError(
+                f"Column 'unrestricted_use' missing from manifest for sample set '{sample_set}'. "
+                "This indicates a data integrity issue in the release manifest."
+            )
+
         sample_set_records_srs = release_manifest_df.loc[
             release_manifest_df["sample_set"] == sample_set, "unrestricted_use"
         ]
@@ -824,12 +831,19 @@ def lookup_study_info(self, sample_set: base_params.sample_set) -> dict:
     def lookup_terms_of_use_info(self, sample_set: base_params.sample_set) -> dict:
         if self._cache_sample_set_to_terms_of_use_info is None:
             df_sample_sets = self._available_sample_sets().set_index("sample_set")
+            expected_cols = [
+                "terms_of_use_expiry_date",
+                "terms_of_use_url",
+                "unrestricted_use",
+            ]
+            missing_cols = [c for c in expected_cols if c not in df_sample_sets.columns]
+            if missing_cols:
+                raise ValueError(
+                    f"Terms-of-use columns missing from manifest: {missing_cols}. "
+                    "This indicates a data integrity issue in the release manifest."
+                )
             self._cache_sample_set_to_terms_of_use_info = df_sample_sets[
-                [
-                    "terms_of_use_expiry_date",
-                    "terms_of_use_url",
-                    "unrestricted_use",
-                ]
+                expected_cols
             ].to_dict(orient="index")
         try:
             return self._cache_sample_set_to_terms_of_use_info[sample_set]

malariagen_data/anoph/h12.py

@@ -630,7 +630,7 @@ def plot_h12_gwss_multi_overlay_track(
         )
 
         # Plot H12.
-        for i, (cohort_label, (x, h12, contig)) in enumerate(res.items()):
+        for i, (cohort_label, (x, h12, contig_idx)) in enumerate(res.items()):
             fig.scatter(
                 x=x,
                 y=h12,

malariagen_data/anoph/snp_data.py

@@ -1872,9 +1872,10 @@ def biallelic_snp_calls(
 
                 # Apply missingness condition.
                 if max_missing_an is not None:
-                    an_missing = (ds_out.sizes["samples"] * ds_out.sizes["ploidy"]) - an
+                    an_total = ds_out.sizes["samples"] * ds_out.sizes["ploidy"]
+                    an_missing = an_total - an
                     if isinstance(max_missing_an, float):
-                        an_missing_frac = an_missing / an
+                        an_missing_frac = an_missing / an_total
                         loc_missing = an_missing_frac <= max_missing_an
                     else:
                         loc_missing = an_missing <= max_missing_an

malariagen_data/anopheles.py

@@ -1066,7 +1066,7 @@ def cohort_diversity_stats(
             cohort_label, cohort_query = cohort
 
         else:
-            raise TypeError(r"invalid cohort parameter: {cohort!r}")
+            raise TypeError(f"invalid cohort parameter: {cohort!r}")
 
         debug("access allele counts")
         ac = self.snp_allele_counts(