sqlframe 3.21.0__tar.gz → 3.21.1__tar.gz

{sqlframe-3.21.0 → sqlframe-3.21.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sqlframe
-Version: 3.21.0
+Version: 3.21.1
 Summary: Turning PySpark Into a Universal DataFrame API
 Home-page: https://github.com/eakmanrq/sqlframe
 Author: Ryan Eakman

{sqlframe-3.21.0 → sqlframe-3.21.1}/docs/duckdb.md

@@ -448,7 +448,6 @@ See something that you would like to see supported? [Open an issue](https://gith
 * [sin](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.sin.html)
 * [size](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.size.html)
 * [skewness](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.skewness.html)
-    * Returned value is different but is still calculating the same thing. Need to investigate difference in calculation
 * [slice](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.slice.html)
 * [sort_array](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.sort_array.html)
 * [soundex](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.soundex.html)

{sqlframe-3.21.0 → sqlframe-3.21.1}/docs/snowflake.md

@@ -487,7 +487,6 @@ See something that you would like to see supported? [Open an issue](https://gith
 * [sinh](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.sinh.html)
 * [size](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.size.html)
 * [skewness](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.skewness.html)
-  * Skewness is calculated differently
 * [slice](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.slice.html)
 * [sort_array](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.sort_array.html)
 * [soundex](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.soundex.html)

{sqlframe-3.21.0 → sqlframe-3.21.1}/sqlframe/_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '3.21.0'
-__version_tuple__ = version_tuple = (3, 21, 0)
+__version__ = version = '3.21.1'
+__version_tuple__ = version_tuple = (3, 21, 1)

{sqlframe-3.21.0 → sqlframe-3.21.1}/sqlframe/base/dataframe.py

@@ -296,6 +296,12 @@ class BaseDataFrame(t.Generic[SESSION, WRITER, NA, STAT, GROUP_DATA]):
 
     @property
     def columns(self) -> t.List[str]:
+        expression_display_names = self.expression.copy()
+        self._set_display_names(expression_display_names)
+        return expression_display_names.named_selects
+
+    @property
+    def _columns(self) -> t.List[str]:
         return self.expression.named_selects
 
     @property
@@ -611,6 +617,18 @@ class BaseDataFrame(t.Generic[SESSION, WRITER, NA, STAT, GROUP_DATA]):
         }
         self.display_name_mapping.update(zipped)
 
+    def _set_display_names(self, select_expression: exp.Select) -> None:
+        for index, column in enumerate(select_expression.expressions):
+            column_name = quote_preserving_alias_or_name(column)
+            if column_name in self.display_name_mapping:
+                display_name_identifier = exp.to_identifier(
+                    self.display_name_mapping[column_name], quoted=True
+                )
+                display_name_identifier._meta = {"case_sensitive": True, **(column._meta or {})}
+                select_expression.expressions[index] = exp.alias_(
+                    column.unalias(), display_name_identifier, quoted=True
+                )
+
     def _get_expressions(
         self,
         optimize: bool = True,
@@ -631,16 +649,7 @@ class BaseDataFrame(t.Generic[SESSION, WRITER, NA, STAT, GROUP_DATA]):
             select_expression = select_expression.transform(
                 replace_id_value, replacement_mapping
             ).assert_is(exp.Select)
-            for index, column in enumerate(select_expression.expressions):
-                column_name = quote_preserving_alias_or_name(column)
-                if column_name in self.display_name_mapping:
-                    display_name_identifier = exp.to_identifier(
-                        self.display_name_mapping[column_name], quoted=True
-                    )
-                    display_name_identifier._meta = {"case_sensitive": True, **(column._meta or {})}
-                    select_expression.expressions[index] = exp.alias_(
-                        column.unalias(), display_name_identifier, quoted=True
-                    )
+            self._set_display_names(select_expression)
             if optimize:
                 select_expression = t.cast(
                     exp.Select,
@@ -1158,8 +1167,8 @@ class BaseDataFrame(t.Generic[SESSION, WRITER, NA, STAT, GROUP_DATA]):
 
     @operation(Operation.FROM)
     def unionByName(self, other: Self, allowMissingColumns: bool = False) -> Self:
-        l_columns = self.columns
-        r_columns = other.columns
+        l_columns = self._columns
+        r_columns = other._columns
         if not allowMissingColumns:
             l_expressions = l_columns
             r_expressions = l_columns
@@ -1619,9 +1628,9 @@ class BaseDataFrame(t.Generic[SESSION, WRITER, NA, STAT, GROUP_DATA]):
         | 16|  Bob|
         +---+-----+
         """
-        if len(cols) != len(self.columns):
+        if len(cols) != len(self._columns):
             raise ValueError(
-                f"Number of column names does not match number of columns: {len(cols)} != {len(self.columns)}"
+                f"Number of column names does not match number of columns: {len(cols)} != {len(self._columns)}"
             )
         expression = self.expression.copy()
         expression = expression.select(

{sqlframe-3.21.0 → sqlframe-3.21.1}/sqlframe/base/function_alternatives.py

@@ -193,10 +193,6 @@ def factorial_ensure_int(col: ColumnOrName) -> Column:
     return Column.invoke_anonymous_function(col_func(col).cast("integer"), "FACTORIAL")
 
 
-def skewness_from_skew(col: ColumnOrName) -> Column:
-    return Column.invoke_anonymous_function(col, "SKEW")
-
-
 def isnan_using_equal(col: ColumnOrName) -> Column:
     lit = get_func_from_session("lit")
     return Column(

{sqlframe-3.21.0 → sqlframe-3.21.1}/sqlframe/base/functions.py

@@ -486,14 +486,32 @@ def var_pop(col: ColumnOrName) -> Column:
 
 @meta(unsupported_engines=["bigquery", "postgres"])
 def skewness(col: ColumnOrName) -> Column:
-    from sqlframe.base.function_alternatives import skewness_from_skew
-
     session = _get_session()
 
+    func_name = "SKEWNESS"
+
     if session._is_snowflake:
-        return skewness_from_skew(col)
+        func_name = "SKEW"
+
+    if session._is_duckdb or session._is_snowflake:
+        when_func = get_func_from_session("when")
+        count_func = get_func_from_session("count")
+        count_star = count_func("*")
+        lit_func = get_func_from_session("lit")
+        sqrt_func = get_func_from_session("sqrt")
+        col = Column.ensure_col(col)
+        return (
+            when_func(count_star == lit_func(0), lit_func(None))
+            .when(count_star == lit_func(1), lit_func(float("nan")))
+            .when(count_star == lit_func(2), lit_func(0.0))
+            .otherwise(
+                Column.invoke_anonymous_function(col, func_name)
+                * (count_star - lit_func(2))
+                / (sqrt_func(count_star * (count_star - lit_func(1))))
+            )
+        )
 
-    return Column.invoke_anonymous_function(col, "SKEWNESS")
+    return Column.invoke_anonymous_function(col, func_name)
 
 
 @meta(unsupported_engines=["bigquery", "postgres"])

{sqlframe-3.21.0 → sqlframe-3.21.1}/sqlframe/base/util.py

@@ -97,12 +97,8 @@ def get_column_mapping_from_schema_input(
     else:
         value = {x.strip(): None for x in schema}
     return {
-        exp.to_column(k).sql(dialect=dialect): exp.DataType.build(v, dialect=dialect)
-        if v is not None
-        else v
-        for k, v in value.items()
+        k: exp.DataType.build(v, dialect=dialect) if v is not None else v for k, v in value.items()
     }
-    # return {x.strip(): None for x in schema}  # type: ignore
 
 
 def get_tables_from_expression_with_join(expression: exp.Select) -> t.List[exp.Table]:

{sqlframe-3.21.0 → sqlframe-3.21.1}/sqlframe.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sqlframe
-Version: 3.21.0
+Version: 3.21.1
 Summary: Turning PySpark Into a Universal DataFrame API
 Home-page: https://github.com/eakmanrq/sqlframe
 Author: Ryan Eakman

{sqlframe-3.21.0 → sqlframe-3.21.1}/tests/integration/engines/test_engine_dataframe.py

@@ -109,6 +109,7 @@ def test_show_from_create_version_1(get_session: t.Callable[[], _BaseSession], c
 +-----+-----+
 """.strip()
     )
+    assert df.columns == ["foo", "BAR"]
 
 
 # https://github.com/eakmanrq/sqlframe/issues/294
@@ -133,8 +134,10 @@ def test_show_from_create_version_2(get_session: t.Callable[[], _BaseSession], c
 +---+-----+
 """.strip()
     )
+    assert df.columns == ["a", "BAR"]
 
 
+# https://github.com/eakmanrq/sqlframe/issues/291
 def test_show_from_create_with_space(get_session: t.Callable[[], _BaseSession], capsys):
     session = get_session()
     df = session.createDataFrame(
@@ -154,3 +157,31 @@ def test_show_from_create_with_space(get_session: t.Callable[[], _BaseSession],
 +--------+
 """.strip()
     )
+
+
+# https://github.com/eakmanrq/sqlframe/issues/291
+def test_show_from_create_with_space_with_schema(get_session: t.Callable[[], _BaseSession], capsys):
+    session = get_session()
+    data = {"an tan": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
+
+    df = session.createDataFrame([*zip(*data.values())], schema=[*data.keys()])
+    df.show()
+    captured = capsys.readouterr()
+    assert (
+        captured.out.strip()
+        == """
++--------+---+-----+
+| an tan | b |  z  |
++--------+---+-----+
+|   1    | 4 | 7.0 |
+|   3    | 4 | 8.0 |
+|   2    | 6 | 9.0 |
++--------+---+-----+
+    """.strip()
+    )
+    assert df.columns == ["an tan", "b", "z"]
+    assert df.collect() == [
+        Row(**{"an tan": 1, "b": 4, "z": 7.0}),
+        Row(**{"an tan": 3, "b": 4, "z": 8.0}),
+        Row(**{"an tan": 2, "b": 6, "z": 9.0}),
+    ]

{sqlframe-3.21.0 → sqlframe-3.21.1}/tests/integration/engines/test_int_functions.py

@@ -690,11 +690,7 @@ def test_var_pop(get_session_and_func):
 def test_skewness(get_session_and_func):
     session, skewness = get_session_and_func("skewness")
     df = session.createDataFrame([[1], [1], [2]], ["c"])
-    if isinstance(session, (DuckDBSession, SnowflakeSession)):
-        # DuckDB calculates skewness differently than spark
-        assert round(df.select(skewness("c")).first()[0], 4) == 1.7321
-    else:
-        assert math.isclose(df.select(skewness(df.c)).first()[0], 0.7071067811865475)
+    assert math.isclose(df.select(skewness(df.c)).first()[0], 0.7071067811865475, rel_tol=1e-5)
 
 
 def test_kurtosis(get_session_and_func):

{sqlframe-3.21.0 → sqlframe-3.21.1}/tests/unit/standalone/test_dataframe.py

@@ -51,7 +51,7 @@ def test_persist_storagelevel(standalone_employee: StandaloneDataFrame, compare_
 
 def test_with_column_duplicate_alias(standalone_employee: StandaloneDataFrame):
     df = standalone_employee.withColumn("fName", F.col("age").cast("string"))
-    assert df.columns == ["employee_id", "fname", "lname", "age", "store_id"]
+    assert df.columns == ["employee_id", "fName", "lname", "age", "store_id"]
     # Make sure that the new columns is added with an alias to `fname`
     assert (
         df.sql(pretty=False)