NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI · #805 most downloaded on PyPI
Always know what to expect from your data.
Last release 9 days ago
25 Sep 2026
Ships on a steady schedule
a new release about every 2 weeks
Nearly every release is documented
notes for 60 of the last 60 stable releases
11 versions withdrawn
withdrawn after publishing
9 years old
369 releases · first in 2017
[FEATURE] Implement CompoundColumnsUnique metric for SqlAlchemyExecutionEngine
One column per quarter.
[FEATURE] GREAT-3439 extended SlackNotificationsAction for slack app tokens (#3440) (Thanks @psheets)
[FEATURE] Create ExpectationValidationGraph class to Maintain Relationship Between Expectation and Metrics and Use it to Associate Exceptions to Expec
Sep 16, 2021 2 files
Sep 16, 2021 2 files
[FEATURE] Rendered Data Doc JSONs can be uploaded and retrieved from GE Cloud
Sep 2, 2021 2 files
Sep 2, 2021 2 files
[FEATURE] Enable GCS DataConnector integration with PandasExecutionEngine
GCS DataConnector integration with PandasExecutionEngine (#3264)InferredAssetGCSDataConnector (#3284)docs (#3312)ConfiguredAssetAzureDataConnector (#3204)GCSDataConnectors (#3301)[FEATURE] Implement Spark Decorators and Helpers; Demonstrate on MulticolumnSumEqual Metric
ConfiguredAssetGCSDataConnector (#3247)[BUGFIX] Fix deprecation warning for importing from collections (#3228) (thanks @ismaildawoodjee)
PandasExecutionEngine to accept Azure DataConnectors (#3214)[FEATURE] Implement ColumnPairValuesInSet metric for PandasExecutionEngine (#3240) [FEATURE] Implement "expect_column_pair_values_A_to_be_greater_than
[FEATURE] Implement ColumnPairValuesInSet metric for PandasExecutionEngine (#3240) [FEATURE] Implement "expect_column_pair_values_A_to_be_greater_than_B" (for PandasExecutionEngine) (#3234) [FEATURE] Implementation of ColumnPairValuesAGreaterThanB metric for PandasExecutionEngine (#3227) [BUGFIX] Wrap optional azure imports in data_connector setup (#3244) [BUGFIX] Change 'to_json_dict' to 'get_json_dict' in Checkpoint (#3231)
[FEATURE] Accept row_condition (with condition_parser) and ignore_row_if parameters for expect_multicolumn_sum_to_equal
conda as installation option in README (#3196) (Thanks @rpanai)test_yaml_config() (#3206)credentials YAML key support for DataConnectors (#3173)[FEATURE] Enable BigQuery tests for Azure CI/CD
--v3-api suite edit to proceed without selecting DataConnectors (#3165)RuntimeBatchRequest is passed to SimpleCheckpoint with RuntimeDataConnector (#3152).csv.gz files (#2695) (Thanks @luke321321)Jul 30, 2021 2 files
Jul 30, 2021 2 files
Jul 22, 2021 2 files
Jul 22, 2021 2 files
[BUGFIX] added expectation_config to ExpectationValidationResult when exception is raised (#2659) (thanks @peterdhansen))
Jul 9, 2021 2 files
Jul 9, 2021 2 files
[DOCS] correct errors and reference complete example for custom expectations (thanks @jdimatteo)
Changelog:
Jun 23, 2021 2 files
Jun 23, 2021 2 files
This release fixes a packaging error that prevented the CLI commands nested under suite from working in the V3 API.
This release fixes a packaging error that prevented the CLI commands nested under suite from working in the V3 API.
[ENHANCEMENT] Improve support for quantiles calculation in Athena
batch_spec_passthrough in configDataConnector.build_batch_spec to use batch_spec_passthrough in configConfiguredAssetSqlDataConnector.build_batch_spec and ConfiguredAssetFilePathDataConnector.build_batch_spec to properly process Asset.batch_spec_passthroughSqlAlchemyExecutionEngine.get_batch_data_and_markers to handle create_temp_table in RuntimeQueryBatchSpecdatasource new notebook for improved data asset inferencecheckpoint new notebookcreate_temp_table info[BREAKING-EXPERIMENTAL] The batch_data attribute of BatchRequest has been removed. To pass in in-memory dataframes at runtime, the new RuntimeDataConn
batch_data attribute of BatchRequest has been removed. To pass in in-memory dataframes at runtime, the new RuntimeDataConnector should be usedRuntimeDataConnector must now be passed Batch Requests of type RuntimeBatchRequestPartitionDefinitionSubset class has been removed - the parent class IDDict is used in its placepartition_request was renamed data_connector_query. The related PartitionRequest class has been removed - the parent class IDDict is used in its placepartition_definition was renamed batch_identifiers. The related PartitionDefinition class has been removed - the parent class IDDict is used in its placePartitionQuery class has been renamed to BatchFilterbatch_identifiers key on DataConnectorQuery (formerly PartitionRequest) has been changed to batch_filter_parametersRuntimeBatchRequest class, which can be used alongside RuntimeDataConnector to specify batches at runtime with either an in-memory dataframe, path (filesystem or s3), or sql queryRuntimeQueryBatchSpec classdata_connector_query contained limit or index #2617[ENHANCEMENT] CLI docs list command implemented for v3 api #2612
docs list command implemented for v3 api #2612docs build command implemented for v3 api #2614docs clean command implemented for v3 api #2615init command implemented for v3 api #2626store list command implemented for v3 api #2627[FEATURE] Added support for references to secrets stores for AWS Secrets Manager, GCP Secret Manager and Azure Key Vault in great_expectations.yml pro
great_expectations.yml project config file (Thanks @Cedric-Magnan!)[MAINTENANCE] Remove deprecated automerge config #2492
[ENHANCEMENT] Improve support for median calculation in Athena (Thanks @kuhnen!) #2521
suite scaffold to work with the UserConfigurableProfiler #2519[FEATURE] Added EmailAction as a new Validation Action (Thanks @Cedric-Magnan!) #2479
[FEATURE] Add "table.head" metric
expect_column_unique_value_count_to_be_between renderer bug (duplicate "Distinct (%)") #2455. Issue #2423moto version < 2.0.0 #2470expect_compound_columns_to_be_unique ExpectationConfig added #2471 Issue #2464[ENHANCEMENT] Optimize tests #2421
markown_text.j2 jinja template #2422suite edit and suite scaffold notebook renderers to output functional validation cells #2432[FEATURE] Add TupleAzureBlobStoreBackend (thanks @syahdeini) #1975
[FEATURE] New implementation of Checkpoints that uses dedicated CheckpointStore (based on the new ConfigurationStore mechanism) #2311, #2338
[BUGFIX] Fix Local variable 'temp_table_schema_name' might be referenced before assignment bug in sqlalchemy_dataset.py #2302
[ENHANCEMENT] Skip checks when great_expectations package did not change #2287
[FEATURE] Add MicrosoftTeamsNotificationAction (Thanks @Antoninj!)
contrib package #2264[MAINTENANCE] Removed mentions of show_cta_footer and added deprecation notes in usage stats #2190. Issue #2120
[ENHANCEMENT] Updated the BigQuery Integration to create a view instead of a table (thanks @alessandrolacorte!) #2082.
DataContext.get_batch() method supports both 0.13 and 0.12 style call arguments[ENHANCEMENT] Support avro format in Spark datasource (thanks @ryanaustincarlson!) #2122
int columns returned incorrect result[ENHANCEMENT] Improved data docs performance by ~30x for large projects and ~4x for smaller projects by changing instantiation of Jinja environment #2
INTRODUCING THE NEW MODULAR EXPECTATIONS API (Experimental): this release introduces a new way to create expectation logic in its own class, making it
Expectation and MetricProvider classes now work together to validate data and consolidate logic for all backends by function. See the how-to guides in our documentation for more information on how to use the new API.parse_strings_as_datetimes and allow_cross_type_comparisons flags in expectations. Expectation Suites that use the flags will need to be updated to use the new Modular Expectations. In general, simply removing the flag will produce correct behavior; if you still want the exact same semantics, you should ensure your raw data already has typed datetime objects.[BUGFIX] Update requirements.txt for ruamel.yaml to >=0.16 - #2048 (thanks @mmetzger!) [BUGFIX] Added option to return scalar instead of list from que
[BUGFIX] Update requirements.txt for ruamel.yaml to >=0.16 - #2048 (thanks @mmetzger!) [BUGFIX] Added option to return scalar instead of list from query store #2060 [BUGFIX] Add missing markdown_content_block_container #2063 [BUGFIX] Fixed a divided by zero error for checkpoints on empty expectation suites #2064 [BUGFIX] Updated sort to correctly return partial unexpected results when expect_column_values_to_be_of_type has more than one unexpected type #2074 [BUGFIX] Resolve Data Docs resource identifier issues to speed up UpdateDataDocs action #2078 [DOCS] Updated contribution changelog location #2051 (thanks @shapiroj18!) [DOCS] Adding Airflow operator and Astrononomer deploy guides #2070 [DOCS] Missing image link to bigquery logo #2071 (thanks @nelsonauner!)
[BUGFIX] Fixed the import of s3fs to use the optional import pattern - issue #2053 [DOCS] Updated the title styling and added a Discuss comment articl
[BUGFIX] Fixed the import of s3fs to use the optional import pattern - issue #2053 [DOCS] Updated the title styling and added a Discuss comment article for the OpsgenieAlertAction how-to guide
[FEATURE] Add OpsgenieAlertAction #2012 (thanks @miike!) [FEATURE] Add S3SubdirReaderBatchKwargsGenerator #2001 (thanks @noklam) [ENHANCEMENT] Snowfla
[FEATURE] Add OpsgenieAlertAction #2012 (thanks @miike!)
[FEATURE] Add S3SubdirReaderBatchKwargsGenerator #2001 (thanks @noklam)
[ENHANCEMENT] Snowflake uses temp tables by default while still allowing transient tables
[ENHANCEMENT] Enabled use of lowercase table and column names in GE with the use_quoted_name key in batch_kwargs #2023
[BUGFIX] Basic suite builder profiler (suite scaffold) now skips excluded expectations #2037
[BUGFIX] Off-by-one error in linking to static images #2036 (thanks @NimaVaziri!)
[BUGFIX] Improve handling of pandas NA type issue #2029 PR #2039 (thanks @isichei!)
[DOCS] Update Virtual Environment Example #2027 (thanks @shapiroj18!)
[DOCS] Update implemented_expectations.rst (thanks @jdimatteo!)
[DOCS] Update how_to_configure_a_pandas_s3_datasource.rst #2042 (thanks @CarstenFrommhold!)
[ENHANCEMENT] CLI supports s3a:// or gs:// paths for Pandas Datasources (issue #2006) [ENHANCEMENT] Escape $ characters in configuration, support mult
[ENHANCEMENT] CLI supports s3a:// or gs:// paths for Pandas Datasources (issue #2006) [ENHANCEMENT] Escape $ characters in configuration, support multiple substitutions (#2005 & #2015) [BUGFIX] Fixed bug where slack messages cause stacktrace when data docs pages have issue [DOCS] Remove incorrect doc line from PagerdutyAlertAction (Thanks @NiallRees!) [MAINTENANCE] Fix path for how-to guide (Thanks @gauthamzz!)
[BUGFIX] replace black in requirements.txt
[ENHANCEMENT] Implement expect_column_values_to_be_json_parseable in spark (Thanks @mikaylaedwards!) [ENHANCEMENT] Fix boto3 options passing into data
[ENHANCEMENT] Implement expect_column_values_to_be_json_parseable in spark (Thanks @mikaylaedwards!) [ENHANCEMENT] Fix boto3 options passing into datasource correctly (Thanks @noklam!) [ENHANCEMENT] Add .pkl to list of recognized extensions (Thanks @KPLauritzen!) [BUGFIX] Query batch kwargs support for Athena backend (issue #1964) [BUGFIX] Skip config substitution if key is "password" (issue #1927) [BUGFIX] fix site_names functionality and add site_names param to get_docs_sites_urls (issue #1991) [BUGFIX] Always render expectation suites in data docs unless passing a specific ExpectationSuiteIdentifier in resource_identifiers (issue #1944) [BUGFIX] remove black from requirements.txt [BUGFIX] docs build cli: fix --yes argument (Thanks @varunbpatil!) [DOCS] Update docstring for SubdirReaderBatchKwargsGenerator (Thanks @KPLauritzen!) [DOCS] Fix broken link in README.md (Thanks @eyaltrabelsi!) [DOCS] Clarifications on several docs (Thanks all!!)
[FEATURE] Add PagerdutyAlertAction (Thanks @NiallRees!)
Sep 28, 2020 2 files
Sep 28, 2020 2 files
[ENHANCEMENT] Update schema for anonymized expectation types to avoid large key domain
[FEATURE] Add expect_column_pair_cramers_phi_value_to_be_less_than expectation to PandasDatasource to check for the independence of two columns by com
expect_column_pair_cramers_phi_value_to_be_less_than expectation to PandasDatasource to check for the independence of two columns by computing their Cramers Phi (thanks @mlondschien)!expect_column_pair_values_to_be_in_set to Spark (thanks @mikaylaedwards)! expect_multicolumn_sum_to_equal for pandas` and Spark`` (thanks @chipmyersjr)![BREAKING] This release includes a breaking change that *only* affects users who directly call add_expectation, remove_expectation, or find_expectatio…
add_expectation, remove_expectation, or find_expectations. (Most users do not use these APIs but add Expectations by stating them directly on Datasets). Those methods have been updated to take an ExpectationConfiguration object and match_type object. The change provides more flexibility in determining which expectations should be modified and allows us provide substantially improved support for two major features that we have frequently heard requested: conditional Expectations and more flexible multi-column custom expectations. See expectation_suite_operations and migrating_versions docs sections for more information.expect_column_values_to_be_increasing to Spark (thanks @mikaylaedwards)!expect_column_values_to_be_decreasing to Spark (thanks @mikaylaedwards)!skip_and_clean_missing flag to DefaultSiteIndexBuilder.build (default True). If True, when an index page is being built and an existing HTML page does not have corresponding source data (i.e. an expectation suite or validation result was removed from source store), the HTML page is automatically deleted and will not appear in the index. This ensures that the expectations store and validations store are the source of truth for Data Docs.config_variables.yml when not at the top levelsuite new[ENHANCEMENT] Deprecate magic "profiling" run_name for Data Docs profiling results rendering
[FEATURE] Customizable "Suite Edit" generated notebooks
allow_update=False to disallow[ENHANCEMENT] Improve CLI error handling.
[FEATURE] Auto-install Python DB packages. If the required packages for a DB library are not installed, GE will offer the user to install them, withou
.feather file support to PandasDatasourcecolorama init to support terminal color on Windows[FEATURE] Add support for expect_column_values_to_match_regex_list exception for Spark backend
[BUGIFX] Fixed an error that crashed the CLI when called in an environment with neither SQLAlchemy nor google.auth installed
[BUGIFX] Fixed an error that crashed the CLI when called in an environment with neither SQLAlchemy nor google.auth installed
[ENHANCEMENT] Removed the misleading scary "Site doesn't exist or is inaccessible" message that the CLI displayed before building Data Docs for the fi
[FEATURE] Add support for expect_volumn_values_to_match_json_schema exception for Spark backend (thanks @chipmyersjr!)
[BUGFIX] Fixed bug that was caused by comparison between timezone aware and non-aware datetimes
Removed deprecated cli tap command
run_id is now typed using the new RunIdentifier class, which consists of a run_time and
run_name. Existing projects that have Expectation Suite Validation Results must be migrated.
See Upgrading to 0.11 for instructions.ValidationMetric and ValidationMetricIdentifier objects now have a data_asset_name attribute.
Existing projects with evaluation parameter stores that have database backends must be migrated.
See Upgrading to 0.11 for instructions.ValidationOperator.run now returns an instance of new type, ValidationOperatorResult (instead of a
dictionary). If your code uses output from Validation Operators, it must be updated.data_asset_name is now added to batch_kwargs by batch_kwargs_generators (if available) and surfaced in Data Docsgenerator_asset parameters to data_asset_namegreat_expectations.yml(BREAKING) run_id is now typed using the new RunIdentifier class, which consists of a run_time and run_name. Existing projects that have Expectation S
run_id is now typed using the new RunIdentifier class, which consists of a run_time and run_name. Existing projects that have Expectation Suite Validation Results must be migrated. See migrating versions for details.ValidationMetric and ValidationMetricIdentifier objects now have a data_asset_name attribute. Existing projects with evaluation parameter stores that have database backends must be migrated. See migrating versions for details.data_asset_name is now added to batch_kwargs by batch_kwargs_generators (if available) and surfaced in Data Docsgenerator_asset parameters to data_asset_nameYour coding agent can read these notes before it upgrades. Set up the MCP server →