From 88be17a691bca098533f2d3657a7f0be2f16a919 Mon Sep 17 00:00:00 2001 From: AlenkaF Date: Tue, 18 Mar 2025 11:01:09 +0100 Subject: [PATCH 1/8] Remove deprecated parquet_2_0 --- .../parquet/arrow/arrow_reader_writer_test.cc | 19 ++++++------------- cpp/src/parquet/metadata.cc | 3 --- cpp/src/parquet/type_fwd.h | 6 ------ python/pyarrow/_parquet.pxd | 1 - python/pyarrow/_parquet.pyx | 8 -------- .../pyarrow/tests/parquet/test_data_types.py | 2 +- python/pyarrow/tests/parquet/test_pandas.py | 2 +- r/R/enums.R | 2 +- r/R/parquet.R | 11 +---------- r/tests/testthat/test-parquet.R | 9 --------- 10 files changed, 10 insertions(+), 53 deletions(-) diff --git a/cpp/src/parquet/arrow/arrow_reader_writer_test.cc b/cpp/src/parquet/arrow/arrow_reader_writer_test.cc index f80ab83c8607..0cc5f948c7dd 100644 --- a/cpp/src/parquet/arrow/arrow_reader_writer_test.cc +++ b/cpp/src/parquet/arrow/arrow_reader_writer_test.cc @@ -1264,7 +1264,7 @@ TEST_F(TestInt96ParquetIO, ReadIntoTimestamp) { using TestUInt32ParquetIO = TestParquetIO<::arrow::UInt32Type>; -TEST_F(TestUInt32ParquetIO, Parquet_2_0_Compatibility) { +TEST_F(TestUInt32ParquetIO, Parquet_2_6_Compatibility) { // This also tests max_definition_level = 1 std::shared_ptr values; @@ -2055,10 +2055,6 @@ TEST(TestArrowReadWrite, ParquetVersionTimestampDifferences) { .version(ParquetVersion::PARQUET_1_0) ->build(); ARROW_SUPPRESS_DEPRECATION_WARNING - auto parquet_version_2_0_properties = ::parquet::WriterProperties::Builder() - .version(ParquetVersion::PARQUET_2_0) - ->build(); - ARROW_UNSUPPRESS_DEPRECATION_WARNING auto parquet_version_2_4_properties = ::parquet::WriterProperties::Builder() .version(ParquetVersion::PARQUET_2_4) ->build(); @@ -2066,8 +2062,8 @@ TEST(TestArrowReadWrite, ParquetVersionTimestampDifferences) { .version(ParquetVersion::PARQUET_2_6) ->build(); const std::vector> all_properties = { - parquet_version_1_properties, parquet_version_2_0_properties, - parquet_version_2_4_properties, parquet_version_2_6_properties}; + parquet_version_1_properties, parquet_version_2_4_properties, + parquet_version_2_6_properties}; { // Using Parquet version 1.0 and 2.4 defaults, seconds should be coerced to @@ -2081,13 +2077,11 @@ TEST(TestArrowReadWrite, ParquetVersionTimestampDifferences) { parquet_version_2_4_properties)); } { - // Using Parquet version 2.0 and 2.6 defaults, seconds should be coerced to + // Using Parquet version 2.6 defaults, seconds should be coerced to // milliseconds and nanoseconds should be retained auto expected_schema = schema({field("ts:s", t_ms), field("ts:ms", t_ms), field("ts:us", t_us), field("ts:ns", t_ns)}); auto expected_table = Table::Make(expected_schema, {a_ms, a_ms, a_us, a_ns}); - ASSERT_NO_FATAL_FAILURE(CheckConfiguredRoundtrip(input_table, expected_table, - parquet_version_2_0_properties)); ASSERT_NO_FATAL_FAILURE(CheckConfiguredRoundtrip(input_table, expected_table, parquet_version_2_6_properties)); } @@ -2133,9 +2127,8 @@ TEST(TestArrowReadWrite, ParquetVersionTimestampDifferences) { CreateOutputStream(), input_table->num_rows(), properties, arrow_coerce_to_nanos_properties)); } - // Using Parquet versions "2.0" and 2.6, coercing to (int64) nanoseconds is allowed - for (const auto& properties : - {parquet_version_2_0_properties, parquet_version_2_6_properties}) { + // Using Parquet version 2.6, coercing to (int64) nanoseconds is allowed + for (const auto& properties : {parquet_version_2_6_properties}) { ARROW_SCOPED_TRACE("format = ", ParquetVersionToString(properties->version())); auto expected_schema = schema({field("ts:s", t_ns), field("ts:ms", t_ns), field("ts:us", t_ns), field("ts:ns", t_ns)}); diff --git a/cpp/src/parquet/metadata.cc b/cpp/src/parquet/metadata.cc index 9b53da021f52..398ff761bd6c 100644 --- a/cpp/src/parquet/metadata.cc +++ b/cpp/src/parquet/metadata.cc @@ -76,9 +76,6 @@ std::string ParquetVersionToString(ParquetVersion::type ver) { case ParquetVersion::PARQUET_1_0: return "1.0"; ARROW_SUPPRESS_DEPRECATION_WARNING - case ParquetVersion::PARQUET_2_0: - return "pseudo-2.0"; - ARROW_UNSUPPRESS_DEPRECATION_WARNING case ParquetVersion::PARQUET_2_4: return "2.4"; case ParquetVersion::PARQUET_2_6: diff --git a/cpp/src/parquet/type_fwd.h b/cpp/src/parquet/type_fwd.h index cda0dc5a77e1..c2e902c41fb7 100644 --- a/cpp/src/parquet/type_fwd.h +++ b/cpp/src/parquet/type_fwd.h @@ -38,12 +38,6 @@ struct ParquetVersion { /// corresponding converted type. PARQUET_1_0, - /// DEPRECATED: Enable Parquet format 2.6 features - /// - /// This misleadingly named enum value is roughly similar to PARQUET_2_6. - PARQUET_2_0 ARROW_DEPRECATED_ENUM_VALUE("use PARQUET_2_4 or PARQUET_2_6 " - "for fine-grained feature selection"), - /// Enable Parquet format 2.4 and earlier features when writing /// /// This enables UINT32 as well as logical types which don't have diff --git a/python/pyarrow/_parquet.pxd b/python/pyarrow/_parquet.pxd index 1e3c89e4e729..e6de9712f833 100644 --- a/python/pyarrow/_parquet.pxd +++ b/python/pyarrow/_parquet.pxd @@ -134,7 +134,6 @@ cdef extern from "parquet/api/schema.h" namespace "parquet" nogil: enum ParquetVersion" parquet::ParquetVersion::type": ParquetVersion_V1" parquet::ParquetVersion::PARQUET_1_0" - ParquetVersion_V2_0" parquet::ParquetVersion::PARQUET_2_0" ParquetVersion_V2_4" parquet::ParquetVersion::PARQUET_2_4" ParquetVersion_V2_6" parquet::ParquetVersion::PARQUET_2_6" diff --git a/python/pyarrow/_parquet.pyx b/python/pyarrow/_parquet.pyx index ecc4e5709176..f47d5da51c99 100644 --- a/python/pyarrow/_parquet.pyx +++ b/python/pyarrow/_parquet.pyx @@ -991,8 +991,6 @@ cdef class FileMetaData(_Weakrefable): cdef ParquetVersion version = self._metadata.version() if version == ParquetVersion_V1: return '1.0' - elif version == ParquetVersion_V2_0: - return 'pseudo-2.0' elif version == ParquetVersion_V2_4: return '2.4' elif version == ParquetVersion_V2_6: @@ -1888,12 +1886,6 @@ cdef shared_ptr[WriterProperties] _create_writer_properties( if version is not None: if version == "1.0": props.version(ParquetVersion_V1) - elif version in ("2.0", "pseudo-2.0"): - warnings.warn( - "Parquet format '2.0' pseudo version is deprecated, use " - "'2.4' or '2.6' for fine-grained feature selection", - FutureWarning, stacklevel=2) - props.version(ParquetVersion_V2_0) elif version == "2.4": props.version(ParquetVersion_V2_4) elif version == "2.6": diff --git a/python/pyarrow/tests/parquet/test_data_types.py b/python/pyarrow/tests/parquet/test_data_types.py index 1428f8023977..855d5952b913 100644 --- a/python/pyarrow/tests/parquet/test_data_types.py +++ b/python/pyarrow/tests/parquet/test_data_types.py @@ -58,7 +58,7 @@ @pytest.mark.pandas @pytest.mark.parametrize('chunk_size', [None, 1000]) -def test_parquet_2_0_roundtrip(tempdir, chunk_size): +def test_parquet_2_6_roundtrip(tempdir, chunk_size): df = alltypes_sample(size=10000, categorical=True) filename = tempdir / 'pandas_roundtrip.parquet' diff --git a/python/pyarrow/tests/parquet/test_pandas.py b/python/pyarrow/tests/parquet/test_pandas.py index 2ea2f46873ae..11655c87d266 100644 --- a/python/pyarrow/tests/parquet/test_pandas.py +++ b/python/pyarrow/tests/parquet/test_pandas.py @@ -121,7 +121,7 @@ def test_pandas_parquet_column_multiindex(tempdir): @pytest.mark.pandas -def test_pandas_parquet_2_0_roundtrip_read_pandas_no_index_written(tempdir): +def test_pandas_parquet_2_roundtrip_read_pandas_no_index_written(tempdir): df = alltypes_sample(size=10000) filename = tempdir / 'pandas_roundtrip.parquet' diff --git a/r/R/enums.R b/r/R/enums.R index 7dd5f2f85881..98995b2a2e70 100644 --- a/r/R/enums.R +++ b/r/R/enums.R @@ -129,7 +129,7 @@ FileType <- enum("FileType", #' @export #' @rdname enums ParquetVersionType <- enum("ParquetVersionType", - PARQUET_1_0 = 0L, PARQUET_2_0 = 1L, PARQUET_2_4 = 2L, PARQUET_2_6 = 3L + PARQUET_1_0 = 0L, PARQUET_2_4 = 2L, PARQUET_2_6 = 3L ) #' @export diff --git a/r/R/parquet.R b/r/R/parquet.R index 88ce1c77128f..2c709fe1b1c9 100644 --- a/r/R/parquet.R +++ b/r/R/parquet.R @@ -232,7 +232,6 @@ ParquetArrowWriterProperties$create <- function(use_deprecated_int96_timestamps valid_parquet_version <- c( "1.0" = ParquetVersionType$PARQUET_1_0, - "2.0" = ParquetVersionType$PARQUET_2_0, "2.4" = ParquetVersionType$PARQUET_2_4, "2.6" = ParquetVersionType$PARQUET_2_6, "latest" = ParquetVersionType$PARQUET_2_6 @@ -252,15 +251,7 @@ make_valid_parquet_version <- function(version, valid_versions = valid_parquet_v call. = FALSE ) } - out <- valid_versions[[arg_match(version, values = names(valid_versions))]] - - if (identical(out, ParquetVersionType$PARQUET_2_0)) { - warning( - 'Parquet format version "2.0" is deprecated. Use "2.4" or "2.6" to select format features.', - call. = FALSE - ) - } - out + valid_versions[[arg_match(version, values = names(valid_versions))]] } #' @title ParquetWriterProperties class diff --git a/r/tests/testthat/test-parquet.R b/r/tests/testthat/test-parquet.R index edca48e92d45..e769aa9ea05f 100644 --- a/r/tests/testthat/test-parquet.R +++ b/r/tests/testthat/test-parquet.R @@ -134,12 +134,6 @@ test_that("make_valid_parquet_version()", { make_valid_parquet_version("1.0"), ParquetVersionType$PARQUET_1_0 ) - expect_deprecated( - expect_equal( - make_valid_parquet_version("2.0"), - ParquetVersionType$PARQUET_2_0 - ) - ) expect_equal( make_valid_parquet_version("2.4"), ParquetVersionType$PARQUET_2_4 @@ -154,9 +148,6 @@ test_that("make_valid_parquet_version()", { ) expect_equal(make_valid_parquet_version(1), ParquetVersionType$PARQUET_1_0) - expect_deprecated( - expect_equal(make_valid_parquet_version(2), ParquetVersionType$PARQUET_2_0) - ) expect_equal(make_valid_parquet_version(1.0), ParquetVersionType$PARQUET_1_0) expect_equal(make_valid_parquet_version(2.4), ParquetVersionType$PARQUET_2_4) }) From ab85561cf3ec0a403703c5980e07429206e2cd40 Mon Sep 17 00:00:00 2001 From: AlenkaF Date: Tue, 18 Mar 2025 15:01:51 +0100 Subject: [PATCH 2/8] Fix test_write_metadata and update parquet.R docstrings --- python/pyarrow/tests/parquet/test_metadata.py | 2 +- r/R/parquet.R | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/python/pyarrow/tests/parquet/test_metadata.py b/python/pyarrow/tests/parquet/test_metadata.py index cf17f830f20e..b3340d93e4d1 100644 --- a/python/pyarrow/tests/parquet/test_metadata.py +++ b/python/pyarrow/tests/parquet/test_metadata.py @@ -554,7 +554,7 @@ def test_write_metadata(tempdir): assert b'ARROW:schema' not in schema_as_arrow.metadata # pass through writer keyword arguments - for version in ["1.0", "2.0", "2.4", "2.6"]: + for version in ["1.0", "2.4", "2.6"]: pq.write_metadata(schema, path, version=version) parquet_meta = pq.read_metadata(path) # The version is stored as a single integer in the Parquet metadata, diff --git a/r/R/parquet.R b/r/R/parquet.R index 2c709fe1b1c9..91ddfc63a297 100644 --- a/r/R/parquet.R +++ b/r/R/parquet.R @@ -98,8 +98,8 @@ read_parquet <- function(file, #' the number of columns and number of rows), though if the data has fewer #' than 250 million cells (rows x cols), then the total number of rows is #' used. -#' @param version parquet version: "1.0", "2.0" (deprecated), "2.4" (default), -#' "2.6", or "latest" (currently equivalent to 2.6). Numeric values are +#' @param version parquet version: "1.0", "2.4" (default), "2.6", or +#' "latest" (currently equivalent to 2.6). Numeric values are #' coerced to character. #' @param compression compression algorithm. Default "snappy". See details. #' @param compression_level compression level. Meaning depends on compression From 5e9a456ce8a54b630210cd40275efe17d234759b Mon Sep 17 00:00:00 2001 From: AlenkaF Date: Tue, 18 Mar 2025 15:14:23 +0100 Subject: [PATCH 3/8] Remove two tests for parquet 2.0 specifically and update write_parquet.Rd --- python/pyarrow/tests/parquet/test_pandas.py | 12 ------------ python/pyarrow/tests/test_pandas.py | 14 -------------- r/man/write_parquet.Rd | 2 +- 3 files changed, 1 insertion(+), 27 deletions(-) diff --git a/python/pyarrow/tests/parquet/test_pandas.py b/python/pyarrow/tests/parquet/test_pandas.py index 11655c87d266..e7144e8e78d3 100644 --- a/python/pyarrow/tests/parquet/test_pandas.py +++ b/python/pyarrow/tests/parquet/test_pandas.py @@ -269,18 +269,6 @@ def test_pandas_parquet_configuration_options(tempdir): tm.assert_frame_equal(df, df_read) -@pytest.mark.pandas -@pytest.mark.filterwarnings("ignore:Parquet format '2.0':FutureWarning") -def test_spark_flavor_preserves_pandas_metadata(): - df = _test_dataframe(size=100) - df.index = np.arange(0, 10 * len(df), 10) - df.index.name = 'foo' - - result = _roundtrip_pandas_dataframe(df, {'version': '2.0', - 'flavor': 'spark'}) - tm.assert_frame_equal(result, df) - - @pytest.mark.pandas def test_index_column_name_duplicate(tempdir): data = { diff --git a/python/pyarrow/tests/test_pandas.py b/python/pyarrow/tests/test_pandas.py index 4ad04c9ad1ec..ad55fc206420 100644 --- a/python/pyarrow/tests/test_pandas.py +++ b/python/pyarrow/tests/test_pandas.py @@ -4891,20 +4891,6 @@ def make_df_with_timestamps(): return df -@pytest.mark.parquet -@pytest.mark.filterwarnings("ignore:Parquet format '2.0':FutureWarning") -def test_timestamp_as_object_parquet(tempdir): - # Timestamps can be stored as Parquet and reloaded into Pandas with no loss - # of information if the timestamp_as_object option is True. - df = make_df_with_timestamps() - table = pa.Table.from_pandas(df) - filename = tempdir / "timestamps_from_pandas.parquet" - pq.write_table(table, filename, version="2.0") - result = pq.read_table(filename) - df2 = result.to_pandas(timestamp_as_object=True) - tm.assert_frame_equal(df, df2) - - def test_timestamp_as_object_out_of_range(): # Out of range timestamps can be converted Arrow and reloaded into Pandas # with no loss of information if the timestamp_as_object option is True. diff --git a/r/man/write_parquet.Rd b/r/man/write_parquet.Rd index 954c692dad2f..ad708b12e74c 100644 --- a/r/man/write_parquet.Rd +++ b/r/man/write_parquet.Rd @@ -32,7 +32,7 @@ the number of columns and number of rows), though if the data has fewer than 250 million cells (rows x cols), then the total number of rows is used.} -\item{version}{parquet version: "1.0", "2.0" (deprecated), "2.4" (default), +\item{version}{parquet version: "1.0", "2.4" (default), "2.6", or "latest" (currently equivalent to 2.6). Numeric values are coerced to character.} From dd686cea0f31d96c463395b4dbed9dd400cc2b2c Mon Sep 17 00:00:00 2001 From: AlenkaF Date: Wed, 19 Mar 2025 08:17:51 +0100 Subject: [PATCH 4/8] Small docstring change to nudge GitHub to process updates --- r/man/write_parquet.Rd | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/r/man/write_parquet.Rd b/r/man/write_parquet.Rd index ad708b12e74c..859006571a09 100644 --- a/r/man/write_parquet.Rd +++ b/r/man/write_parquet.Rd @@ -32,8 +32,8 @@ the number of columns and number of rows), though if the data has fewer than 250 million cells (rows x cols), then the total number of rows is used.} -\item{version}{parquet version: "1.0", "2.4" (default), -"2.6", or "latest" (currently equivalent to 2.6). Numeric values are +\item{version}{parquet version: "1.0", "2.4" (default), "2.6", or +"latest" (currently equivalent to 2.6). Numeric values are coerced to character.} \item{compression}{compression algorithm. Default "snappy". See details.} From 378e847626e6a64a79725bfbf9d10e91d55bc035 Mon Sep 17 00:00:00 2001 From: AlenkaF Date: Tue, 25 Mar 2025 09:34:39 +0100 Subject: [PATCH 5/8] Fix linter errors --- python/pyarrow/tests/parquet/test_pandas.py | 3 +-- python/pyarrow/tests/test_pandas.py | 5 ----- 2 files changed, 1 insertion(+), 7 deletions(-) diff --git a/python/pyarrow/tests/parquet/test_pandas.py b/python/pyarrow/tests/parquet/test_pandas.py index e7144e8e78d3..a5490e680c5a 100644 --- a/python/pyarrow/tests/parquet/test_pandas.py +++ b/python/pyarrow/tests/parquet/test_pandas.py @@ -41,8 +41,7 @@ import pandas as pd import pandas.testing as tm - from pyarrow.tests.parquet.common import (_roundtrip_pandas_dataframe, - alltypes_sample) + from pyarrow.tests.parquet.common import alltypes_sample except ImportError: pd = tm = None diff --git a/python/pyarrow/tests/test_pandas.py b/python/pyarrow/tests/test_pandas.py index ad55fc206420..50dc00dd5df4 100644 --- a/python/pyarrow/tests/test_pandas.py +++ b/python/pyarrow/tests/test_pandas.py @@ -47,11 +47,6 @@ from pyarrow.vendored.version import Version import pyarrow as pa -try: - from pyarrow import parquet as pq -except ImportError: - pass - try: import pandas as pd import pandas.testing as tm From 4173150963d22cba6c2f585e3911f52c4ee304c5 Mon Sep 17 00:00:00 2001 From: AlenkaF Date: Tue, 25 Mar 2025 15:37:04 +0100 Subject: [PATCH 6/8] Revert "Remove two tests for parquet 2.0 specifically and update write_parquet.Rd" This reverts commit 5e9a456ce8a54b630210cd40275efe17d234759b. --- python/pyarrow/tests/parquet/test_pandas.py | 12 ++++++++++++ python/pyarrow/tests/test_pandas.py | 14 ++++++++++++++ 2 files changed, 26 insertions(+) diff --git a/python/pyarrow/tests/parquet/test_pandas.py b/python/pyarrow/tests/parquet/test_pandas.py index a5490e680c5a..646faa0cb62d 100644 --- a/python/pyarrow/tests/parquet/test_pandas.py +++ b/python/pyarrow/tests/parquet/test_pandas.py @@ -268,6 +268,18 @@ def test_pandas_parquet_configuration_options(tempdir): tm.assert_frame_equal(df, df_read) +@pytest.mark.pandas +@pytest.mark.filterwarnings("ignore:Parquet format '2.0':FutureWarning") +def test_spark_flavor_preserves_pandas_metadata(): + df = _test_dataframe(size=100) + df.index = np.arange(0, 10 * len(df), 10) + df.index.name = 'foo' + + result = _roundtrip_pandas_dataframe(df, {'version': '2.0', + 'flavor': 'spark'}) + tm.assert_frame_equal(result, df) + + @pytest.mark.pandas def test_index_column_name_duplicate(tempdir): data = { diff --git a/python/pyarrow/tests/test_pandas.py b/python/pyarrow/tests/test_pandas.py index 50dc00dd5df4..45b3dd5a7be6 100644 --- a/python/pyarrow/tests/test_pandas.py +++ b/python/pyarrow/tests/test_pandas.py @@ -4886,6 +4886,20 @@ def make_df_with_timestamps(): return df +@pytest.mark.parquet +@pytest.mark.filterwarnings("ignore:Parquet format '2.0':FutureWarning") +def test_timestamp_as_object_parquet(tempdir): + # Timestamps can be stored as Parquet and reloaded into Pandas with no loss + # of information if the timestamp_as_object option is True. + df = make_df_with_timestamps() + table = pa.Table.from_pandas(df) + filename = tempdir / "timestamps_from_pandas.parquet" + pq.write_table(table, filename, version="2.0") + result = pq.read_table(filename) + df2 = result.to_pandas(timestamp_as_object=True) + tm.assert_frame_equal(df, df2) + + def test_timestamp_as_object_out_of_range(): # Out of range timestamps can be converted Arrow and reloaded into Pandas # with no loss of information if the timestamp_as_object option is True. From 52114cd5f8133ee032d16555f0c9a46364dbd80b Mon Sep 17 00:00:00 2001 From: AlenkaF Date: Tue, 25 Mar 2025 15:43:04 +0100 Subject: [PATCH 7/8] Fix Python tests to use default parquet writer --- python/pyarrow/tests/parquet/test_pandas.py | 7 +++---- python/pyarrow/tests/test_pandas.py | 8 ++++++-- 2 files changed, 9 insertions(+), 6 deletions(-) diff --git a/python/pyarrow/tests/parquet/test_pandas.py b/python/pyarrow/tests/parquet/test_pandas.py index 646faa0cb62d..58a39c1967ea 100644 --- a/python/pyarrow/tests/parquet/test_pandas.py +++ b/python/pyarrow/tests/parquet/test_pandas.py @@ -41,7 +41,8 @@ import pandas as pd import pandas.testing as tm - from pyarrow.tests.parquet.common import alltypes_sample + from pyarrow.tests.parquet.common import (_roundtrip_pandas_dataframe, + alltypes_sample) except ImportError: pd = tm = None @@ -269,14 +270,12 @@ def test_pandas_parquet_configuration_options(tempdir): @pytest.mark.pandas -@pytest.mark.filterwarnings("ignore:Parquet format '2.0':FutureWarning") def test_spark_flavor_preserves_pandas_metadata(): df = _test_dataframe(size=100) df.index = np.arange(0, 10 * len(df), 10) df.index.name = 'foo' - result = _roundtrip_pandas_dataframe(df, {'version': '2.0', - 'flavor': 'spark'}) + result = _roundtrip_pandas_dataframe(df, {'flavor': 'spark'}) tm.assert_frame_equal(result, df) diff --git a/python/pyarrow/tests/test_pandas.py b/python/pyarrow/tests/test_pandas.py index 45b3dd5a7be6..39757c1f4cb3 100644 --- a/python/pyarrow/tests/test_pandas.py +++ b/python/pyarrow/tests/test_pandas.py @@ -47,6 +47,11 @@ from pyarrow.vendored.version import Version import pyarrow as pa +try: + from pyarrow import parquet as pq +except ImportError: + pass + try: import pandas as pd import pandas.testing as tm @@ -4887,14 +4892,13 @@ def make_df_with_timestamps(): @pytest.mark.parquet -@pytest.mark.filterwarnings("ignore:Parquet format '2.0':FutureWarning") def test_timestamp_as_object_parquet(tempdir): # Timestamps can be stored as Parquet and reloaded into Pandas with no loss # of information if the timestamp_as_object option is True. df = make_df_with_timestamps() table = pa.Table.from_pandas(df) filename = tempdir / "timestamps_from_pandas.parquet" - pq.write_table(table, filename, version="2.0") + pq.write_table(table, filename) result = pq.read_table(filename) df2 = result.to_pandas(timestamp_as_object=True) tm.assert_frame_equal(df, df2) From 2e0387b510e10141e3d7343ffbce0f6e09530105 Mon Sep 17 00:00:00 2001 From: Bryce Mecum Date: Wed, 26 Mar 2025 11:13:46 -0700 Subject: [PATCH 8/8] Run devtools::document() --- r/man/enums.Rd | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/r/man/enums.Rd b/r/man/enums.Rd index dd0ca944b8c6..e2f50dfc86dc 100644 --- a/r/man/enums.Rd +++ b/r/man/enums.Rd @@ -36,7 +36,7 @@ An object of class \code{Compression::type} (inherits from \code{arrow-enum}) of An object of class \code{FileType} (inherits from \code{arrow-enum}) of length 4. -An object of class \code{ParquetVersionType} (inherits from \code{arrow-enum}) of length 4. +An object of class \code{ParquetVersionType} (inherits from \code{arrow-enum}) of length 3. An object of class \code{MetadataVersion} (inherits from \code{arrow-enum}) of length 5.