lenskit.operations

Top-level LensKit functions for recommender operations.

Functions

recommend(pipeline, query[, n, items, component, profiler])	Generate recommendations for a user or query. This calls the specified
score(pipeline, query, items, *[, component, profiler])	Score items with respect to a user or query. This calls the specified
predict(pipeline, query, items, *[, component, profiler])	Predict ratings for items. This is exactly like score(), except it

Module Contents

lenskit.operations.recommend(pipeline, query, n=None, items=None, *, component='recommender', profiler=None)

Generate recommendations for a user or query. This calls the specified pipeline component (the 'recommender' by default) and returns the resulting item list.

Stability:

Full (see Stability Levels).

Parameters:

• pipeline (lenskit.pipeline.Pipeline) – The pipeline to run.

• query (lenskit.data.QueryInput) – The user ID or other query data for the recommendation.

• n (int | None) – The number of items to recommend.

• items (lenskit.data.ItemList | lenskit.data.types.IDSequence | None) – The candidate items, or None to use the pipeline’s default candidate selector.

• component (str) – The name of the component implementing the recommender.

• profiler (lenskit.pipeline.PipelineProfiler | None) – A profiler for profiling this pipeline run.

Returns:

The recommended items as an ordered item list.

Return type:

lenskit.data.ItemList

lenskit.operations.score(pipeline, query, items, *, component='scorer', profiler=None)

Score items with respect to a user or query. This calls the specified pipeline component (the 'scorer' by default) and returns the resulting item list.

Stability:

Full (see Stability Levels).

Parameters:

• pipeline (lenskit.pipeline.Pipeline) – The pipeline to run.

• query (lenskit.data.QueryInput) – The user ID or other query data for the recommendation.

• items (lenskit.data.ItemList | lenskit.data.types.IDSequence) – The candidate items, or None to use the pipeline’s default candidate selector.

• component (str) – The name of the component implementing the scorer.

• profiler (lenskit.pipeline.PipelineProfiler | None)

Returns:

The items with scores in the score field.

Return type:

lenskit.data.ItemList

lenskit.operations.predict(pipeline, query, items, *, component='rating-predictor', profiler=None)

Predict ratings for items. This is exactly like score(), except it defaults to the 'rating-predictor' component. In a standard pipeline, the rating predictor may have additional configuration such as fallbacks or transformations to ensure every item is scored and the scores are valid rating predictions; the scorer typically returns raw scores.

Stability:

Full (see Stability Levels).

Parameters:

• pipeline (lenskit.pipeline.Pipeline)

• query (lenskit.data.QueryInput)

• items (lenskit.data.ItemList | lenskit.data.types.IDSequence)

• component (str)

• profiler (lenskit.pipeline.PipelineProfiler | None)

Return type:

lenskit.data.ItemList

Exported Aliases

class lenskit.operations.ItemList

Re-exported alias for lenskit.data.ItemList.

lenskit.operations.QueryInput

Re-exported alias for lenskit.data.QueryInput.

lenskit.operations.IDSequence

Re-exported alias for lenskit.data.types.IDSequence.

class lenskit.operations.Pipeline

Re-exported alias for lenskit.pipeline.Pipeline.

class lenskit.operations.PipelineProfiler

Re-exported alias for lenskit.pipeline.PipelineProfiler.