Evaluation

Overview

Evaluations store the test results from a single execution of the test suite. They are constructed either using Fetchers or manually by specifying a Pandas dataframe and evaluation ID.

Processing

Evaluations can be processed by using the process() method. Its only parameter is a list of Processors that will be used to transform the evaluation one-at-a-time.

Evaluation.process(pipeline: List[Processor]) → Evaluation

Executes each processor in the pipeline and returns a new Evaluation.

Parameters:pipeline (a list of Processors to process the Evaluation in order) – Returns:an Evaluation instance that was processed by the pipeline

Extracting the internal dataframe

Evaluations store the test results internally using a Pandas dataframe. You can retrieve this dataframe by using the get_df() method. This can be used to manipulate the underlying data without needing to use the built-in FTPVL processors.

Note

get_df() returns a defensive copy of the dataframe, so any mutations to the returned dataframe will not be reflected in the original Evaluation. Instead, you must instantiate a new Evaluation by passing in the dataframe and evaluation ID.

Evaluation.get_df() → pandas.core.frame.DataFrame

Returns a copy of the Pandas DataFrame that represents the evaluation

Example

>>> eval1 = Evaluation(pd.DataFrame([{"a": 1, "b": 2}, {"a": 4, "b": 5}]), eval_id=1)
>>> df = eval1.get_df() # extract Pandas dataframe
>>> df
    a   b
0   1   2
1   4   5
>>> df["c"] = [3, 6] # add a new column
>>> eval2 = Evaluation(df, eval_id=eval1.get_eval_id()) # create new evaluation
>>> eval2.get_df()
    a   b   c
0   1   2   3
1   4   5   6