Fetchers

Overview

Fetchers serve as a convenient way to ingest information from a source and create an Evaluation. Sources can either be local or over the network.

Fetching from Hydra

You can fetch from Hydra by using the HydraFetcher fetcher. Its functionality is explained in the Fetching Data from Hydra section of the First Steps guide.

Fetching from a JSON dataframe

You can also fetch from a properly-formatted JSON dataframe by using the JSONFetcher fetcher and specifying the path to the JSON file in the parameter. This is useful if some pre-processing needs to be performed, or the test runner is able to output a dataframe as a build artifact.

Most commonly, this fetcher can import JSON-encoded dataframes if they are exported from a separate Pandas dataframe. Learn about exporting as a JSON file from Pandas.

Getting the Evaluation

To get the Evaluation object from the Fetcher, call the get_evaluation() method.

Fetcher.get_evaluation() → ftpvl.evaluation.Evaluation

Returns an Evaluation that represents the fetched data.