[SPARK-28041][PYTHON] Increase minimum supported Pandas to 0.23.2

## What changes were proposed in this pull request?

This increases the minimum supported version of Pandas to 0.23.2. Using a lower version will raise an error `Pandas >= 0.23.2 must be installed; however, your version was 0.XX`. Also, a workaround for using pyarrow with Pandas 0.19.2 was removed.

## How was this patch tested?

Existing Tests

Closes #24867 from BryanCutler/pyspark-increase-min-pandas-SPARK-28041.

Authored-by: Bryan Cutler <cutlerb@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>

docs/sql-migration-guide-upgrade.md

@@ -23,6 +23,10 @@ license: |
 {:toc}
 
 ## Upgrading From Spark SQL 2.4 to 3.0
+  - Since Spark 3.0, PySpark requires a Pandas version of 0.23.2 or higher to use Pandas related functionality, such as `toPandas`, `createDataFrame` from Pandas DataFrame, etc.
+
+  - Since Spark 3.0, PySpark requires a PyArrow version of 0.12.1 or higher to use PyArrow related functionality, such as `pandas_udf`, `toPandas` and `createDataFrame` with "spark.sql.execution.arrow.enabled=true", etc.
+
   - In Spark version 2.4 and earlier, SQL queries such as `FROM <table>` or `FROM <table> UNION ALL FROM <table>` are supported by accident. In hive-style `FROM <table> SELECT <expr>`, the `SELECT` clause is not negligible. Neither Hive nor Presto support this syntax. Therefore we will treat these queries as invalid since Spark 3.0.
 
   - Since Spark 3.0, the Dataset and DataFrame API `unionAll` is not deprecated any more. It is an alias for `union`.

python/pyspark/serializers.py

@@ -297,8 +297,6 @@ def create_array(s, t):
             # Ensure timestamp series are in expected form for Spark internal representation
             if t is not None and pa.types.is_timestamp(t):
                 s = _check_series_convert_timestamps_internal(s.fillna(0), self._timezone)
-                # TODO: need cast after Arrow conversion, ns values cause error with pandas 0.19.2
-                return pa.Array.from_pandas(s, mask=mask).cast(t, safe=False)
 
             try:
                 array = pa.Array.from_pandas(s, mask=mask, type=t, safe=self._safecheck)

python/pyspark/sql/tests/test_arrow.py

@@ -268,10 +268,10 @@ def test_createDataFrame_with_schema(self):
     def test_createDataFrame_with_incorrect_schema(self):
         pdf = self.create_pandas_data_frame()
         fields = list(self.schema)
-        fields[0], fields[7] = fields[7], fields[0]  # swap str with timestamp
+        fields[0], fields[1] = fields[1], fields[0]  # swap str with int
         wrong_schema = StructType(fields)
         with QuietTest(self.sc):
-            with self.assertRaisesRegexp(Exception, ".*cast.*[s|S]tring.*timestamp.*"):
+            with self.assertRaisesRegexp(Exception, "integer.*required.*got.*str"):
                 self.spark.createDataFrame(pdf, schema=wrong_schema)
 
     def test_createDataFrame_with_names(self):

python/pyspark/sql/utils.py

@@ -131,7 +131,7 @@ def require_minimum_pandas_version():
     """ Raise ImportError if minimum version of Pandas is not installed
     """
     # TODO(HyukjinKwon): Relocate and deduplicate the version specification.
-    minimum_pandas_version = "0.19.2"
+    minimum_pandas_version = "0.23.2"
 
     from distutils.version import LooseVersion
     try:

python/setup.py

@@ -105,7 +105,7 @@ def _supports_symlinks():
 # If you are changing the versions here, please also change ./python/pyspark/sql/utils.py
 # For Arrow, you should also check ./pom.xml and ensure there are no breaking changes in the
 # binary format protocol with the Java version, see ARROW_HOME/format/* for specifications.
-_minimum_pandas_version = "0.19.2"
+_minimum_pandas_version = "0.23.2"
 _minimum_pyarrow_version = "0.12.1"
 
 try: