fixup(data-pipeline): sort pext regions on region start in pipeline

broadinstitute · Nov 14, 2024 · b7a8d67 · b7a8d67
1 parent 08a0636
commit b7a8d67
Showing 1 changed file with 9 additions and 4 deletions.
diff --git a/data-pipeline/src/data_pipeline/data_types/pext.py b/data-pipeline/src/data_pipeline/data_types/pext.py
@@ -189,10 +189,15 @@ def reshape_pext_data_to_tissue_array(pext_struct_path):
         ds = ds.annotate(flags=hl.empty_array(hl.tstr))
 
     ds = ds.annotate(
-        regions=ds.regions.map(
-            lambda region: region.annotate(
-                tissues=hl.array([hl.struct(tissue=tissue, value=region.tissues[tissue]) for tissue in region.tissues])
-            )
+        regions=hl.sorted(
+            ds.regions.map(
+                lambda region: region.annotate(
+                    tissues=hl.array(
+                        [hl.struct(tissue=tissue, value=region.tissues[tissue]) for tissue in region.tissues]
+                    )
+                )
+            ),
+            key=lambda region: region.start,
         )
     )
     return ds