def assess_speed(df_kind): rounds = 100 input_df = input_test_data()[df_kind][0] orig_config = anr._FAST_CHECK_DF_SERIALIZABLE try: _bench(rounds, input_df, fast=False) _bench(rounds, input_df, fast=True) finally: anr._FAST_CHECK_DF_SERIALIZABLE = orig_config
def test_dataframe_confirm_fast_check_compatibility(input_df): orig_config = anr.FAST_CHECK_DF_SERIALIZABLE try: input_df = input_test_data()[input_df][0] anr.set_fast_check_df_serializable(True) with_fast_check = df_serializer.can_convert_to_records_without_objects(input_df, 'symA') anr.set_fast_check_df_serializable(False) without_fast_check = df_serializer.can_convert_to_records_without_objects(input_df, 'symA') assert with_fast_check == without_fast_check finally: anr.FAST_CHECK_DF_SERIALIZABLE = orig_config
import pytest import arctic.serialization.numpy_records as anr from tests.unit.serialization.serialization_test_data import _mixed_test_data as input_test_data df_serializer = anr.DataFrameSerializer() @pytest.mark.parametrize("input_df", input_test_data().keys()) def test_dataframe_confirm_fast_check_compatibility(input_df): orig_config = anr.FAST_CHECK_DF_SERIALIZABLE try: input_df = input_test_data()[input_df][0] anr.set_fast_check_df_serializable(True) with_fast_check = df_serializer.can_convert_to_records_without_objects( input_df, 'symA') anr.set_fast_check_df_serializable(False) without_fast_check = df_serializer.can_convert_to_records_without_objects( input_df, 'symA') assert with_fast_check == without_fast_check finally: anr.FAST_CHECK_DF_SERIALIZABLE = orig_config