[CI][Python] test_arrow_cogrouped_map and test_arrow_grouped_map from pyspark fail on integration job

[raulcd]

Describe the bug, including details regarding any error messages, version, and platform.

As seen here:
#44981 (comment)
When I tried to run "pyspark.sql.tests.arrow.test_arrow_grouped_map" and "pyspark.sql.tests.arrow.test_arrow_cogrouped_map" they fail due to missing pandas:

 Traceback (most recent call last):
  File "/spark/python/pyspark/sql/tests/arrow/test_arrow_grouped_map.py", line 264, in test_self_join
    df2 = df.groupby("k").applyInArrow(arrow_func, schema="x long, y long")
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/spark/python/pyspark/sql/pandas/group_ops.py", line 809, in applyInArrow
    udf = pandas_udf(
          ^^^^^^^^^^^
  File "/spark/python/pyspark/sql/pandas/functions.py", line 372, in pandas_udf
    require_minimum_pandas_version()
  File "/spark/python/pyspark/sql/pandas/utils.py", line 43, in require_minimum_pandas_version
    raise PySparkImportError(
pyspark.errors.exceptions.base.PySparkImportError: [PACKAGE_NOT_INSTALLED] Pandas >= 2.0.0 must be installed; however, it was not found.

Those tests were never executed in the past but might be worth to include them on the job as they are arrow related.

Component(s)

Continuous Integration, Python