ExperimentResult#

class pymc_marketing.mmm.experiment.ExperimentResult(result, experiment_type)[source]#

Wrapper around a CausalPy experiment result.

Provides convenience methods to access the causal effect estimates and convert them into the lift test DataFrame format expected by MMM.add_lift_test_measurements().

Parameters:

resultcausalpy experiment instance: The CausalPy experiment result (e.g. InterruptedTimeSeries).
experiment_typeExperimentType: The type of experiment that was run.

Attributes:

resultcausalpy experiment instance: The underlying CausalPy result object.
experiment_typeExperimentType: The experiment type.

Examples

from pymc_marketing.mmm.experiment import ExperimentResult, ExperimentType

experiment_result = ExperimentResult(
    result=causalpy_result,
    experiment_type=ExperimentType.ITS,
)
experiment_result.summary()
df_lift = experiment_result.to_lift_test(channel="tv", x=1000.0, delta_x=200.0)

Methods

`ExperimentResult.__init__`(result, ...)
`ExperimentResult.effect_summary`(**kwargs)	Get a structured effect summary.
`ExperimentResult.plot`(**kwargs)	Plot the experiment results.
`ExperimentResult.summary`()	Print a summary of the experiment results.
`ExperimentResult.to_lift_test`(channel, x, ...)	Convert experiment results to a lift test DataFrame.

Attributes

idata

Access the ArviZ InferenceData object.