Support optional inputs in model signatures

[apurva-koti]

Related Issues/PRs

#xxx

What changes are proposed in this pull request?

Add an optional boolean parameter to ColSpec to specify whether the column in question is required for model inference or can be omitted. Updates code accordingly:

• Updates pyfunc schema checks to check against missing required columns. Missing optional columns will be ignored, but provided optional columns will still be type checked.
• Defaults optional to False for backwards compatibility.
• Prevents argument autofill in pyfunc models returned as spark UDFs as added here Enable spark_udf to use column names from model signature by default #4236

Example usage (pyfunc):

input_schema = Schema(
    [
        ColSpec("double", "a"),
        ColSpec("double", "b"),
        ColSpec("string", "c", optional=True),
        ColSpec("long", "d", optional=True),
    ]
)
signature = ModelSignature(inputs=input_schema)
data = {"a": [1.0], "b": [1.0]}
data_2 = {"a": [1.0], "b": [1.0], "d": [2]}

for data in [
    data,
    data_2,
]:
    pd_data = pd.DataFrame(data)
    check = _enforce_schema(pd_data, signature.inputs) #passes

Example (spark udf):

test_signature = {
    "inputs": '[{"name": "a", "type": "long"}, {"name": "b", "type": "long"}, {"name" : "c", "type": "long", "optional": "True"}]',
}
signature = ModelSignature.from_dict(test_signature)
...
udf = mlflow.pyfunc.spark_udf(...)

data = spark.createDataFrame(
    pd.DataFrame(columns=["a", "b"], data={"a": [1, 2], "b": [2, 3]})
)

res = data.withColumn("response", udf(*data.columns)) #

How is this patch tested?

• Existing unit/integration tests
• New unit/integration tests
• Manual tests (describe details, including test results, below)

Does this PR change the documentation?

• No. You can skip the rest of this section.
• Yes. Make sure the changed pages / sections render correctly in the documentation preview.

Release Notes

Optional input columns can now be specified in model signatures. These columns can be omitted from input dataframes at prediction time.

Is this a user-facing change?

• No. You can skip the rest of this section.
• Yes. Give a description of this change to be included in the release notes for MLflow users.

Optional input columns can now be specified in model signatures. These columns can be omitted from input dataframes at prediction time.

What component(s), interfaces, languages, and integrations does this PR affect?

Components

• area/artifacts: Artifact stores and artifact logging
• area/build: Build and test infrastructure for MLflow
• area/docs: MLflow documentation pages
• area/examples: Example code
• area/model-registry: Model Registry service, APIs, and the fluent client calls for Model Registry
• area/models: MLmodel format, model serialization/deserialization, flavors
• area/recipes: Recipes, Recipe APIs, Recipe configs, Recipe Templates
• area/projects: MLproject format, project running backends
• area/scoring: MLflow Model server, model deployment tools, Spark UDFs
• area/server-infra: MLflow Tracking server backend
• area/tracking: Tracking Service, tracking client APIs, autologging

Interface

• area/uiux: Front-end, user experience, plotting, JavaScript, JavaScript dev server
• area/docker: Docker use across MLflow's components, such as MLflow Projects and MLflow Models
• area/sqlalchemy: Use of SQLAlchemy in the Tracking Service or Model Registry
• area/windows: Windows support

Language

• language/r: R APIs and clients
• language/java: Java APIs and clients
• language/new: Proposals for new client languages

Integrations

• integrations/azure: Azure and Azure ML integrations
• integrations/sagemaker: SageMaker integrations
• integrations/databricks: Databricks integrations

How should the PR be classified in the release notes? Choose one:

• rn/breaking-change - The PR will be mentioned in the "Breaking Changes" section
• rn/none - No description will be included. The PR will be mentioned only by the PR number in the "Small Bugfixes and Documentation Updates" section
• rn/feature - A new user-facing feature worth mentioning in the release notes
• rn/bug-fix - A user-facing bug fix worth mentioning in the release notes
• rn/documentation - A user-facing documentation change worth mentioning in the release notes

[mlflow-automation]

Documentation preview for 96030fb will be available here when this CircleCI job completes successfully.

More info

• Ignore this comment if this PR does not change the documentation.
• It takes a few minutes for the preview to be available.
• The preview is updated when a new commit is pushed to this PR.
• This comment was created by https://github.com/mlflow/mlflow/actions/runs/5085462433.

[WeichenXu123]

Could you update PR description to attach an example code that shows the case your PR supports ?

[apurva-koti]

@WeichenXu123 PR not ready yet. I'll have all that in when requesting review

[apurva-koti]

Automatic parameter filling relies on the model signature to determine what the columns will be when the UDF is applied to the dataframe.
With optional columns, we can neither:

• include one or all of them, as that would implicitly require those columns to exist in the dataframe, raising an error within pyspark
• exclude one or all of them, as they would then not be selected from the dataframe at all

Given this, it seems reasonable to prevent this convenience for this case. Users can still manually pass in the list of columns to udf as follows:

test_signature = {
    "inputs": '[{"name": "a", "type": "long"}, {"name": "b", "type": "long"}, {"name" : "c", "type": "long", "optional": "True"}]',
}
signature = ModelSignature.from_dict(test_signature)
...
udf = mlflow.pyfunc.spark_udf(...)

data = spark.createDataFrame(
    pd.DataFrame(columns=["a", "b"], data={"a": [1, 2], "b": [2, 3]})
)

res = data.withColumn("response", udf(*data.columns)) # <-- calling udf() would throw an exception.

[WeichenXu123]

Make sense!

[WeichenXu123]

LGTM