diff --git a/spark_frame/data_diff/diff_result.py b/spark_frame/data_diff/diff_result.py
index f0556da..22e3db2 100644
--- a/spark_frame/data_diff/diff_result.py
+++ b/spark_frame/data_diff/diff_result.py
@@ -432,11 +432,15 @@ def _compute_top_per_col_state_df(self, diff_df: DataFrame) -> DataFrame:
             "diff.right_value",
             "diff.sample_ids",
         ).where(exists_in_left_or_right)
-        window = Window.partitionBy("column_name", "state").orderBy(
-            f.col("nb").desc(),
-            f.col("left_value"),
-            f.col("right_value"),
-        ).rowsBetween(Window.unboundedPreceding, Window.currentRow)
+        window = (
+            Window.partitionBy("column_name", "state")
+            .orderBy(
+                f.col("nb").desc(),
+                f.col("left_value"),
+                f.col("right_value"),
+            )
+            .rowsBetween(Window.unboundedPreceding, Window.currentRow)
+        )
         df_2 = (
             df_1.groupBy("column_name", "state", "left_value", "right_value")
             .agg(f.count(f.lit(1)).alias("nb"), f.first("sample_ids").alias("sample_ids"))
diff --git a/spark_frame/graph_impl/ascending_forest_traversal.py b/spark_frame/graph_impl/ascending_forest_traversal.py
index 47575db..287284d 100644
--- a/spark_frame/graph_impl/ascending_forest_traversal.py
+++ b/spark_frame/graph_impl/ascending_forest_traversal.py
@@ -456,11 +456,13 @@ def ascending_forest_traversal(
     )
 
     if keep_labels:
-        res_df = res_df.join(input_df, node_id).select(
+        res_df = res_df.join(input_df, node_id, how="inner").select(
             res_df["*"],
             f.struct(*[input_df[quote(col)] for col in input_df.columns]).alias("node"),
         )
-        res_df = res_df.join(input_df.alias("input_df"), res_df[quote(parent_id)] == input_df[quote(node_id)]).select(
+        res_df = res_df.join(
+            input_df.alias("input_df"), res_df[quote(parent_id)] == input_df[quote(node_id)], how="inner",
+        ).select(
             res_df[quote(node_id)],
             res_df[quote(parent_id)],
             res_df["node"],
diff --git a/spark_frame/transformations_impl/analyze_aggs.py b/spark_frame/transformations_impl/analyze_aggs.py
index de3222b..345f749 100644
--- a/spark_frame/transformations_impl/analyze_aggs.py
+++ b/spark_frame/transformations_impl/analyze_aggs.py
@@ -1,39 +1,39 @@
-from pyspark.sql import Column
-from pyspark.sql import functions as f
-from pyspark.sql.types import StructField
-
-
-def _to_string(col: Column) -> Column:
-    return col.cast("STRING")
-
-
-def column_number(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, NOSONAR
-    return f.lit(col_num).alias("column_number")
-
-
-def column_name(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, NOSONAR
-    return f.lit(struct_field.name).alias("column_name")
-
-
-def column_type(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, NOSONAR
-    return f.lit(struct_field.dataType.typeName().upper()).alias("column_type")
-
-
-def count(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, NOSONAR
-    return f.count(f.lit(1)).alias("count")
-
-
-def count_distinct(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, NOSONAR
-    return f.count_distinct(col).alias("count_distinct")
-
-
-def count_null(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, NOSONAR
-    return (f.count(f.lit(1)) - f.count(col)).alias("count_null")
-
-
-def min(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, A001, NOSONAR
-    return _to_string(f.min(col)).alias("min")
-
-
-def max(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001, A001, NOSONAR
-    return _to_string(f.max(col)).alias("max")
+from pyspark.sql import Column
+from pyspark.sql import functions as f
+from pyspark.sql.types import StructField
+
+
+def _to_string(col: Column) -> Column:
+    return col.cast("STRING")
+
+
+def column_number(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,NOSONAR(S1172)
+    return f.lit(col_num).alias("column_number")
+
+
+def column_name(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,NOSONAR(S1172)
+    return f.lit(struct_field.name).alias("column_name")
+
+
+def column_type(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,NOSONAR(S1172)
+    return f.lit(struct_field.dataType.typeName().upper()).alias("column_type")
+
+
+def count(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,NOSONAR(S1172)
+    return f.count(f.lit(1)).alias("count")
+
+
+def count_distinct(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,NOSONAR(S1172)
+    return f.count_distinct(col).alias("count_distinct")
+
+
+def count_null(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,NOSONAR(S1172)
+    return (f.count(f.lit(1)) - f.count(col)).alias("count_null")
+
+
+def min(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,A001,NOSONAR(S1172)
+    return _to_string(f.min(col)).alias("min")
+
+
+def max(col: str, struct_field: StructField, col_num: int) -> Column:  # noqa: ARG001,A001,NOSONAR(S1172)
+    return _to_string(f.max(col)).alias("max")