Merge branch 'master' into GH407_allow_gene_labels

Author: leehart

malariagen_data/anoph/base.py

@@ -327,14 +327,18 @@ def _discover_releases(self) -> Tuple[str, ...]:
         )
         # Note: this matches v3, v3. and v3.1, but not v3001.1
         version_pattern = re.compile(f"^v{self._major_version_number}(\\..*)?$")
+        # To sort the versions numerically, we use a lambda function for the "key" parameter of sorted().
+        # The lambda function splits each version string into a list of its integer parts, using split('.') and int(), e.g. [3, 1],
+        # which sorted() then uses to determine the order, as opposed to the default lexicographic order.
         discovered_releases = tuple(
             sorted(
                 [
                     self._path_to_release(d)
                     for d in sub_dirs
                     if version_pattern.match(d)
                     and self._fs.exists(f"{self._base_path}/{d}/manifest.tsv")
-                ]
+                ],
+                key=lambda v: [int(part) for part in v.split(".")],
             )
         )
         return discovered_releases

malariagen_data/anopheles.py

@@ -1191,6 +1191,11 @@ def _gene_cnv_frequencies(
 
                 freq_cols[f"frq_{coh}"] = np.concatenate([amp_freq_coh, del_freq_coh])
 
+        if len(coh_dict) == 0:
+            raise ValueError(
+                "No cohorts available for the given sample selection parameters and minimum cohort size."
+            )
+
         debug("build a dataframe with the frequency columns")
         df_freqs = pd.DataFrame(freq_cols)