agglovar 0.0.1.dev9__tar.gz → 0.0.1.dev10__tar.gz

{agglovar-0.0.1.dev9/src/agglovar.egg-info → agglovar-0.0.1.dev10}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agglovar
-Version: 0.0.1.dev9
+Version: 0.0.1.dev10
 Summary: Toolkit for fast genomic variant transformations and intersects
 License-Expression: MIT
 Requires-Python: >=3.12

{agglovar-0.0.1.dev9 → agglovar-0.0.1.dev10}/src/agglovar/__init__.py

@@ -1,6 +1,6 @@
 """Agglovar: A toolkit for fast genomic variant transformations and intersects."""
 
-__version__ = '0.0.1.dev9'
+__version__ = '0.0.1.dev10'
 
 __all__ = [
     'align',

{agglovar-0.0.1.dev9 → agglovar-0.0.1.dev10}/src/agglovar/bed/intersect.py

@@ -86,6 +86,17 @@ def as_proportion(
     if name is None or not (name := name.strip()):
         raise ValueError('Name must be a non-empty string')
 
+    if isinstance(df_a, pl.DataFrame):
+        df_a = df_a.lazy()
+
+    if isinstance(df_b, pl.DataFrame):
+        df_b = df_b.lazy()
+
+    col_set_a = set(df_a.collect_schema().names())
+
+    if '_index' not in col_set_a:
+        df_a = df_a.with_row_index('_index')
+
     # Collapse b, index a
     df_b_nr = (
         merge_depth(df_b, 0, col_names_b)
@@ -96,8 +107,6 @@ def as_proportion(
 
     df_a = (
         df_a
-        .drop('_index', strict=False)
-        .with_row_index('_index')
         .filter(
             pl.col(col_names_a.pos).is_not_null(),
             pl.col(col_names_a.end).is_not_null(),
@@ -128,11 +137,12 @@ def as_proportion(
         df_a
         .select(
             '_index',
+            '_index_a_bed_as_prop',
             (col_expr_a.end - col_expr_a.pos).alias('len')
         )
         .join(
             df_join.lazy(),
-            on='_index',
+            on='_index_a_bed_as_prop',
             how='left',
         )
         .select(

{agglovar-0.0.1.dev9 → agglovar-0.0.1.dev10}/src/agglovar/bed/join.py

@@ -46,6 +46,15 @@ class _JoinResources:
         if isinstance(df_b, pl.DataFrame):
             df_b = df_b.lazy()
 
+        cols_a = set(df_a.collect_schema().keys())
+        cols_b = set(df_b.collect_schema().keys())
+
+        if '_index' not in cols_a:
+            df_a = df_a.with_row_index('_index')
+
+        if '_index' not in cols_b:
+            df_b = df_b.with_row_index('_index')
+
         # Set column names
         ref_cols = get_coord_cols('ref')
 
@@ -65,16 +74,18 @@ class _JoinResources:
         # Prepare tables
         df_a = (
             df_a
-            .select(*col_select_a)
-            .drop('_index_a', strict=False)
-            .with_row_index('_index_a')
+            .select(
+                pl.col('_index').alias('_index_a'),
+                *col_select_a
+            )
         )
 
         df_b = (
             df_b
-            .select(*col_select_b)
-            .drop('_index_b', strict=False)
-            .with_row_index('_index_b')
+            .select(
+                pl.col('_index').alias('_index_b'),
+                *col_select_b
+            )
         )
 
         self.df_a = df_a
@@ -322,13 +333,10 @@ def pairwise_join_tree(
 
     for df_batch_b in (
         df_b
-        .drop('_index_b', strict=False)
-        .with_row_index('_index_b')
         .select(pl.col('_index_b'), *col_b.exprs())
         .collect_batches(chunk_size=chunk_size)
     ):
         for index_b, chrom_b, pos_b, end_b in df_batch_b.iter_rows():
-            # print(f'Adding: {chrom_b} - {pos_b} - {end_b}')
             itree[chrom_b].addi(pos_b - distance, end_b + distance, index_b)
 
     # Intersect
@@ -336,8 +344,6 @@ def pairwise_join_tree(
 
     for df_chunk in (
         df_a
-        .drop('_index_a', strict=False)
-        .with_row_index('_index_a')
         .select(pl.col('_index_a'), *col_a.exprs())
         .collect_batches(chunk_size=chunk_size)
     ):

{agglovar-0.0.1.dev9 → agglovar-0.0.1.dev10/src/agglovar.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agglovar
-Version: 0.0.1.dev9
+Version: 0.0.1.dev10
 Summary: Toolkit for fast genomic variant transformations and intersects
 License-Expression: MIT
 Requires-Python: >=3.12