GeezerVison

The World Wide Web, restored to 800 pixels of glory.

GeezerVison is showing github.com — feat: Incorporate substrait ODFVs into ibis-based offline store queries by tokoko · Pull Request #4102 · feast-dev/feast

Conversation

tokoko commented Apr 16, 2024

Copy link

Copy Markdown

Collaborator

What this PR does / why we need it:

This PR changes the offline flow for ODFV execution in ibis-based offline stores (currently only duckdb). Instead of collecting raw offline store output into a pyarrow table and applying a substrait transformation with acero, ibis-based offline stores now get ibis functions and directly extend ibis logical plan, meaning ODFVs will be executed by the offline store engine itself.

The feature necessitated a couple of underlying changes in substrait ODFVs:

substrait ODVS now also store ibis python functions in protos to directly apply these functions in transform_ibis. This can probably be avoided in the future if there's a good substrait to ibis compiler in place, but currently is the simplest solution. Note: there are other problems here as well, substrait plans need to know precisely what input columns they will consume, this is problematic in the offline flow as there's no way to know beforehand which features the user will request in get_historical_features call. To correctly apply ODFVs, one would need to apply each ODFV to a relevant subset of columns and then join all tables back together (kind of like how it's done in the online flow), which is a horrible waste of resources.
Unlike pandas transformations, ibis functions must now return all the columns that are passed as an input to the function (in addition to on demand features) as opposed to returning only the on demand features that they generate. This is to avoid an extra join back to the main table after ibis functions are applied to the output. This way multiple ODFVs can easily be chained, each mutating the offline store output to append the relevant columns.

With this PR, all of the planned ibis/substrait features will now be in place in python sdk. The next steps will be to enable native substrait ODFV execution in non-python sdks as well.

Which issue(s) this PR fixes:

Fixes #3979

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

zilto reacted with heart emoji

All reactions

❤️ 1 reaction


          ibis offline stores execute substrait odfvs natively in ibis

06dcc4d

Signed-off-by: tokoko <[email protected]>

haoxu0 reviewed

Apr 16, 2024

View reviewed changes

haoxu0 left a comment

Copy link

Copy Markdown

Collaborator

There was a problem hiding this comment.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, maybe a bad question, does it mean the substrait transformation is only available in ibis offline store? Is it possible to use it with other offline store? Suppose I have a s3 file sources and want to use the substrait transform directly

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

All reactions

tokoko commented Apr 16, 2024

Copy link

Copy Markdown

Collaborator Author

@HaoXuAI No, it's applicable for all offline stores, the difference is that for non-ibis offline stores the whole dataset will be collected to a single process as an arrow table and transform_arrow method will be used to apply transformation. (This is true for both substrait and pandas transformations)

feast/sdk/python/feast/transformation/substrait_transformation.py

Line 37 in 5c4a9c5

def transform_arrow(self, pa_table: pyarrow.Table) -> pyarrow.Table:

What this PR adds on top of this is that for ibis-based offline stores collect to arrow is no longer necessary and instead of transform_arrow method, newly added transform_ibis will be used to directly apply transformation with the offline store engine.

P.S. duckdb offline store which is for now the the only ibis-based implementation can work with s3 file sources even now.

haoxu0 reacted with thumbs up emoji

All reactions

👍 1 reaction

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

haoxu0 reviewed

Apr 16, 2024

View reviewed changes

Comment thread

sdk/python/feast/transformation/substrait_transformation.py Resolved

Uh oh!

There was an error while loading. Please reload this page.

haoxu0 approved these changes

Apr 17, 2024

View reviewed changes

haoxu0 left a comment

Copy link

Copy Markdown

Collaborator

There was a problem hiding this comment.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lgtm！

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

All reactions

haoxu0 added the ok-to-test label

Apr 17, 2024

haoxu0 merged commit c3a102f into feast-dev:master

Apr 19, 2024

tokoko deleted the duckdb-substrait-odfv branch

April 19, 2024 17:51

franciscojavierarceo pushed a commit that referenced this pull request

May 24, 2024


          chore(release): release 0.38.0

65de0d2

# [0.38.0](v0.37.0...v0.38.0) (2024-05-24)

### Bug Fixes

* Add vector database doc ([#4165](#4165)) ([37f36b6](37f36b6))
* Change checkout action back to v3 from v5 which isn't released yet ([#4147](#4147)) ([9523fff](9523fff))
* Change numpy version <1.25 dependency to <2 in setup.py ([#4085](#4085)) ([2ba71ff](2ba71ff)), closes [#4084](#4084)
* Changed the code the way mysql container is initialized.  ([#4140](#4140)) ([8b5698f](8b5698f)), closes [#4126](#4126)
* Correct nightly install command, move all installs to uv ([#4164](#4164)) ([c86d594](c86d594))
* Default value is not set in Redis connection string using environment variable ([#4136](#4136)) ([95acfb4](95acfb4)), closes [#3669](#3669)
* Get container host addresses from testcontainers (java) ([#4125](#4125)) ([9184dde](9184dde))
* Get rid of empty string `name_alias` during feature view projection deserialization  ([#4116](#4116)) ([65056ce](65056ce))
* Helm chart `feast-feature-server`, improve Service template name ([#4161](#4161)) ([dedc164](dedc164))
* Improve the code related to on-demand-featureview. ([#4203](#4203)) ([d91d7e0](d91d7e0))
* Integration tests for async sdk method ([#4201](#4201)) ([08c44ae](08c44ae))
* Make sure schema is used when calling `get_table_query_string` method for Snowflake datasource ([#4131](#4131)) ([c1579c7](c1579c7))
* Make sure schema is used when generating `from_expression` for Snowflake ([#4177](#4177)) ([5051da7](5051da7))
* Pass native input values to `get_online_features` from feature server ([#4117](#4117)) ([60756cb](60756cb))
* Pass region to S3 client only if set (Java) ([#4151](#4151)) ([b8087f7](b8087f7))
* Pgvector patch ([#4108](#4108)) ([ad45bb4](ad45bb4))
* Update doc ([#4153](#4153)) ([e873636](e873636))
* Update master-only benchmark bucket name due to credential update ([#4183](#4183)) ([e88f1e3](e88f1e3))
* Updating the instructions for quickstart guide. ([#4120](#4120)) ([0c30e96](0c30e96))
* Upgrading the test container so that local tests works with updated d… ([#4155](#4155)) ([93ddb11](93ddb11))

### Features

* Add a Kubernetes Operator for the Feast Feature Server ([#4145](#4145)) ([4a696dc](4a696dc))
* Add delta format to `FileSource`, add support for it in ibis/duckdb ([#4123](#4123)) ([2b6f1d0](2b6f1d0))
* Add materialization support to ibis/duckdb ([#4173](#4173)) ([369ca98](369ca98))
* Add optional private key params to Snowflake config ([#4205](#4205)) ([20f5419](20f5419))
* Add s3 remote storage export for duckdb ([#4195](#4195)) ([6a04c48](6a04c48))
* Adding DatastoreOnlineStore 'database' argument. ([#4180](#4180)) ([e739745](e739745))
* Adding get_online_features_async to feature store sdk ([#4172](#4172)) ([311efc5](311efc5))
* Adding support for dictionary writes to online store  ([#4156](#4156)) ([abfac01](abfac01))
* Elasticsearch vector database ([#4188](#4188)) ([bf99640](bf99640))
* Enable other distance metrics for Vector DB and Update docs ([#4170](#4170)) ([ba9f4ef](ba9f4ef))
* Feast/IKV datetime edgecase errors ([#4211](#4211)) ([bdae562](bdae562))
* Feast/IKV documenation language changes ([#4149](#4149)) ([690a621](690a621))
* Feast/IKV online store contrib plugin integration ([#4068](#4068)) ([f2b4eb9](f2b4eb9))
* Feast/IKV online store documentation ([#4146](#4146)) ([73601e4](73601e4))
* Feast/IKV upgrade client version ([#4200](#4200)) ([0e42150](0e42150))
* Incorporate substrait ODFVs into ibis-based offline store queries ([#4102](#4102)) ([c3a102f](c3a102f))
* Isolate input-dependent calculations in `get_online_features` ([#4041](#4041)) ([2a6edea](2a6edea))
* Make arrow primary interchange for online ODFV execution ([#4143](#4143)) ([3fdb716](3fdb716))
* Move data source validation entrypoint to offline store ([#4197](#4197)) ([a17725d](a17725d))
* Upgrading python version to 3.11, adding support for 3.11 as well. ([#4159](#4159)) ([4b1634f](4b1634f)), closes [#4152](#4152) [#4114](#4114)

### Reverts

* Reverts "fix: Using version args to install the correct feast version" ([#4112](#4112)) ([b66baa4](b66baa4)), closes [#3953](#3953)

franciscojavierarceo pushed a commit that referenced this pull request

May 27, 2024


          chore(release): release 0.38.0

277bad4

# [0.38.0](v0.37.0...v0.38.0) (2024-05-24)

### Bug Fixes

* Add vector database doc ([#4165](#4165)) ([37f36b6](37f36b6))
* Change checkout action back to v3 from v5 which isn't released yet ([#4147](#4147)) ([9523fff](9523fff))
* Change numpy version <1.25 dependency to <2 in setup.py ([#4085](#4085)) ([2ba71ff](2ba71ff)), closes [#4084](#4084)
* Changed the code the way mysql container is initialized.  ([#4140](#4140)) ([8b5698f](8b5698f)), closes [#4126](#4126)
* Correct nightly install command, move all installs to uv ([#4164](#4164)) ([c86d594](c86d594))
* Default value is not set in Redis connection string using environment variable ([#4136](#4136)) ([95acfb4](95acfb4)), closes [#3669](#3669)
* Get container host addresses from testcontainers (java) ([#4125](#4125)) ([9184dde](9184dde))
* Get rid of empty string `name_alias` during feature view projection deserialization  ([#4116](#4116)) ([65056ce](65056ce))
* Helm chart `feast-feature-server`, improve Service template name ([#4161](#4161)) ([dedc164](dedc164))
* Improve the code related to on-demand-featureview. ([#4203](#4203)) ([d91d7e0](d91d7e0))
* Integration tests for async sdk method ([#4201](#4201)) ([08c44ae](08c44ae))
* Make sure schema is used when calling `get_table_query_string` method for Snowflake datasource ([#4131](#4131)) ([c1579c7](c1579c7))
* Make sure schema is used when generating `from_expression` for Snowflake ([#4177](#4177)) ([5051da7](5051da7))
* Pass native input values to `get_online_features` from feature server ([#4117](#4117)) ([60756cb](60756cb))
* Pass region to S3 client only if set (Java) ([#4151](#4151)) ([b8087f7](b8087f7))
* Pgvector patch ([#4108](#4108)) ([ad45bb4](ad45bb4))
* Update doc ([#4153](#4153)) ([e873636](e873636))
* Update master-only benchmark bucket name due to credential update ([#4183](#4183)) ([e88f1e3](e88f1e3))
* Updating the instructions for quickstart guide. ([#4120](#4120)) ([0c30e96](0c30e96))
* Upgrading the test container so that local tests works with updated d… ([#4155](#4155)) ([93ddb11](93ddb11))

### Features

* Add a Kubernetes Operator for the Feast Feature Server ([#4145](#4145)) ([4a696dc](4a696dc))
* Add delta format to `FileSource`, add support for it in ibis/duckdb ([#4123](#4123)) ([2b6f1d0](2b6f1d0))
* Add materialization support to ibis/duckdb ([#4173](#4173)) ([369ca98](369ca98))
* Add optional private key params to Snowflake config ([#4205](#4205)) ([20f5419](20f5419))
* Add s3 remote storage export for duckdb ([#4195](#4195)) ([6a04c48](6a04c48))
* Adding DatastoreOnlineStore 'database' argument. ([#4180](#4180)) ([e739745](e739745))
* Adding get_online_features_async to feature store sdk ([#4172](#4172)) ([311efc5](311efc5))
* Adding support for dictionary writes to online store  ([#4156](#4156)) ([abfac01](abfac01))
* Elasticsearch vector database ([#4188](#4188)) ([bf99640](bf99640))
* Enable other distance metrics for Vector DB and Update docs ([#4170](#4170)) ([ba9f4ef](ba9f4ef))
* Feast/IKV datetime edgecase errors ([#4211](#4211)) ([bdae562](bdae562))
* Feast/IKV documenation language changes ([#4149](#4149)) ([690a621](690a621))
* Feast/IKV online store contrib plugin integration ([#4068](#4068)) ([f2b4eb9](f2b4eb9))
* Feast/IKV online store documentation ([#4146](#4146)) ([73601e4](73601e4))
* Feast/IKV upgrade client version ([#4200](#4200)) ([0e42150](0e42150))
* Incorporate substrait ODFVs into ibis-based offline store queries ([#4102](#4102)) ([c3a102f](c3a102f))
* Isolate input-dependent calculations in `get_online_features` ([#4041](#4041)) ([2a6edea](2a6edea))
* Make arrow primary interchange for online ODFV execution ([#4143](#4143)) ([3fdb716](3fdb716))
* Move data source validation entrypoint to offline store ([#4197](#4197)) ([a17725d](a17725d))
* Upgrading python version to 3.11, adding support for 3.11 as well. ([#4159](#4159)) ([4b1634f](4b1634f)), closes [#4152](#4152) [#4114](#4114)

### Reverts

* Reverts "fix: Using version args to install the correct feast version" ([#4112](#4112)) ([b66baa4](b66baa4)), closes [#3953](#3953)

Reviewers

haoxu0

haoxu0 approved these changes

Uh oh!

There was an error while loading. Please reload this page.

Assignees

No one assigned

Labels

ok-to-test

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Abstract ibis-based Offline Store

Uh oh!

There was an error while loading. Please reload this page.

2 participants

Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.

You are visitor 01402503 — GeezerVison does not store the pages you look up.
Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required.