Module: meridian.analysis.review.results

Data structures for the Model Quality Checks results.

Classes

class BaseCase: Base class for all check cases.

class BaseResultData: Base class for check result data.

class BaselineCases: Cases for the Baseline Check.

class BaselineCheckResult: The immutable result of the Baseline Check.

class BayesianPPPCases: Cases for the Bayesian Posterior Predictive P-value Check.

class BayesianPPPCheckResult: The immutable result of the Bayesian Posterior Predictive P-value Check.

class ChannelResult: Base class for channel-level check results.

class CheckResult: Base class for model-level check results.

class ConvergenceCases: Cases for the Convergence Check.

class ConvergenceCheckResult: The immutable result of the Convergence Check.

class GoodnessOfFitCases: Cases for the Goodness of Fit Check.

class GoodnessOfFitCheckResult: The immutable result of the Goodness of Fit Check.

class ModelCheckCase: Base class for all model-level check cases.

class PriorPosteriorShiftAggregateCases: Cases for Prior-Posterior Shift Check aggregate result.

class PriorPosteriorShiftChannelCases: Cases for Prior-Posterior Shift Check per channel.

class PriorPosteriorShiftChannelResult: The result of Prior-Posterior Shift Check for a single channel.

class PriorPosteriorShiftCheckResult: The immutable result of model-level Prior-Posterior Shift Check.

class ROIConsistencyAggregateCases: Cases for ROI Consistency Check aggregate result.

class ROIConsistencyChannelCases: Cases for ROI Consistency Check per channel.

class ROIConsistencyChannelResult: The immutable result of ROI Consistency Check for a single channel.

class ROIConsistencyCheckResult: The immutable result of model-level ROI Consistency Check.

class ReviewSummary: The final summary of all model quality checks.

class Status: Create a collection of name/value pairs.

NOT_CONVERGED_RECOMMENDATION ('We recommend increasing MCMC iterations or investigating model ' 'misspecification (e.g., priors, multicollinearity) before proceeding.')
NOT_FULLY_CONVERGED_RECOMMENDATION ('Manually inspect the parameters with high R-hat values to determine if the ' 'results are acceptable for your use case, and consider increasing MCMC ' 'iterations or investigating model misspecification.')