API Reference

This page contains the API reference for public objects and function within TrustyAI. See the (example notebooks) for usage guides and tutorials.

trustyai.initializer

Initializing The JVM

init(*args[, path])

Manually initialize the JVM.

trustyai.model

Feature and Output Objects

feature(name, dtype[, value, domain])	Create a Java Feature.
feature_domain(values)	Create a Java FeatureDomain.
output(name, dtype[, value, score])	Create a Java Output.

Data Objects

Dataset()

Wrapper class for TrustyAI Datasets.

Model Classes

Model(fn[, pandas, arrow])

Wrap any Python predictive model.

trustyai.explainers

LIME

LimeExplainer(**kwargs)	"Which features were most important to the results?"
LimeResults(saliencyResults)	Wraps LIME results.

SHAP

SHAPExplainer(background[, link_type])	"By how much did each feature contribute to the outputs?"
BackgroundGenerator(datapoints[, ...])	Generate a background for the SHAP explainer
SHAPResults(saliency_results, background)	Wraps SHAP results.

Counterfactuals

CounterfactualExplainer([steps])	"How do I get the result I want?"
CounterfactualResult(result)	Wraps Counterfactual results.

trustyai.utils

Tyrus(model, inputs, outputs, background, ...)

The TrustyAI Assistant and Dashboard.