Skip to content

[SPARK-41209][PYTHON] Improve PySpark type inference in _merge_type method #38731

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
sadikovi wants to merge 3 commits into from
Closed

[SPARK-41209][PYTHON] Improve PySpark type inference in _merge_type method #38731

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
087d071
update code
sadikovi Nov 21, 2022
119d81f
revert NullType change
sadikovi Nov 21, 2022
07fa05b
fix code style
sadikovi Nov 22, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
46 changes: 32 additions & 14 deletions python/pyspark/sql/tests/test_types.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -235,13 +235,21 @@ def test_infer_schema_not_enough_names(self):
df = self.spark.createDataFrame([["a", "b"]], ["col1"])
self.assertEqual(df.columns, ["col1", "_2"])

def test_infer_schema_fails(self):
with self.assertRaisesRegex(TypeError, "field a"):
self.spark.createDataFrame(
self.spark.sparkContext.parallelize([[1, 1], ["x", 1]]),
schema=["a", "b"],
samplingRatio=0.99,
)
def test_infer_schema_upcast_int_to_string(self):
df = self.spark.createDataFrame(
self.spark.sparkContext.parallelize([[1, 1], ["x", 1]]),
schema=["a", "b"],
samplingRatio=0.99,
)
self.assertEqual([Row(a="1", b=1), Row(a="x", b=1)], df.collect())

def test_infer_schema_upcast_float_to_string(self):
df = self.spark.createDataFrame([[1.33, 1], ["2.1", 1]], schema=["a", "b"])
self.assertEqual([Row(a="1.33", b=1), Row(a="2.1", b=1)], df.collect())

def test_infer_schema_upcast_boolean_to_string(self):
df = self.spark.createDataFrame([[True, 1], ["false", 1]], schema=["a", "b"])
self.assertEqual([Row(a="true", b=1), Row(a="false", b=1)], df.collect())

def test_infer_nested_schema(self):
NestedRow = Row("f1", "f2")
Expand Down Expand Up @@ -316,8 +324,10 @@ def test_infer_array_merge_element_types(self):
self.assertRaises(ValueError, lambda: self.spark.createDataFrame(data3))

# an array with conflicting types should raise an error
# in this case this is ArrayType(StringType) and ArrayType(NullType)
data4 = [ArrayRow([1, "1"], [None])]
self.assertRaises(TypeError, lambda: self.spark.createDataFrame(data4))
with self.assertRaisesRegex(ValueError, "types cannot be determined after inferring"):
self.spark.createDataFrame(data4)

def test_infer_array_element_type_empty(self):
# SPARK-39168: Test inferring array element type from all rows
Expand Down Expand Up @@ -840,8 +850,12 @@ def test_merge_type(self):
_merge_type(MapType(StringType(), LongType()), MapType(StringType(), LongType())),
MapType(StringType(), LongType()),
)
with self.assertRaisesRegex(TypeError, "key of map"):
_merge_type(MapType(StringType(), LongType()), MapType(DoubleType(), LongType()))

self.assertEqual(
_merge_type(MapType(StringType(), LongType()), MapType(DoubleType(), LongType())),
MapType(StringType(), LongType()),
)

with self.assertRaisesRegex(TypeError, "value of map"):
_merge_type(MapType(StringType(), LongType()), MapType(StringType(), DoubleType()))

Expand All @@ -865,11 +879,13 @@ def test_merge_type(self):
),
StructType([StructField("f1", StructType([StructField("f2", LongType())]))]),
)
with self.assertRaisesRegex(TypeError, "field f2 in field f1"):
self.assertEqual(
_merge_type(
StructType([StructField("f1", StructType([StructField("f2", LongType())]))]),
StructType([StructField("f1", StructType([StructField("f2", StringType())]))]),
)
),
StructType([StructField("f1", StructType([StructField("f2", StringType())]))]),
)

self.assertEqual(
_merge_type(
Expand Down Expand Up @@ -937,11 +953,13 @@ def test_merge_type(self):
),
StructType([StructField("f1", ArrayType(MapType(StringType(), LongType())))]),
)
with self.assertRaisesRegex(TypeError, "key of map element in array field f1"):
self.assertEqual(
_merge_type(
StructType([StructField("f1", ArrayType(MapType(StringType(), LongType())))]),
StructType([StructField("f1", ArrayType(MapType(DoubleType(), LongType())))]),
)
),
StructType([StructField("f1", ArrayType(MapType(StringType(), LongType())))]),
)

# test for SPARK-16542
def test_array_types(self):
Expand Down
4 changes: 4 additions & 0 deletions python/pyspark/sql/types.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1648,6 +1648,10 @@ def new_name(n: str) -> str:
return a
elif isinstance(a, TimestampNTZType) and isinstance(b, TimestampType):
return b
elif isinstance(a, AtomicType) and isinstance(b, StringType):
return b
elif isinstance(a, StringType) and isinstance(b, AtomicType):
return a
elif type(a) is not type(b):
# TODO: type cast (such as int -> long)
raise TypeError(new_msg("Can not merge type %s and %s" % (type(a), type(b))))
Expand Down