API Query Guide#
This selection of guides and references covers the basics of querying data using the Arthur Query API Endpoints. The complete reference for the full Arthur API documentation can be found here, and the reference to the Query Endpoint can be found here.
Query Quickstart#
For beginners to the Arthur Query API, see the Query Quickstart for a basic walkthrough on querying data about your model’s performance.
References:#
Endpoint Overview#
The Endpoint Overview covers the structure and parameters of queries and provides more examples to get started.
Model Evaluation Functions#
This reference shows the basic metrics needed to evaluate the performance of models (accuracy, confusion matrix, MSE, etc)
Data Drift#
This reference shows querying data drift over time, which is needed to diagnose which changes in data could explain a change in model performance.
Explainability#
This reference shows how to query explanations for inferences. Whether in terms of features in tabular models, words in NLP models, or regions of pixels in CV models, explainability offers a rough approximation of the how and why behind a model’s inferences.
Aggregation Functions#
This reference shows how to apply aggregators like an average (avg
) or a maximum (max
) to a query.
Transformation Functions#
This reference shows the transformations available within an Arthur query (numerical operations, binning variables, grouping by time resolution, etc)
Composing Functions#
This reference shows composing functions to fit complex transformations into a single query
Subqueries#
This reference shows how to embed queries within queries.
Grouped Inference Queries#
This guide shows how to query information pertaining to individual instances of data over time.