Module: tfp.experimental.stats

Statistical functions.

Classes

class RunningCentralMoments: Holds metadata for and computes running central moments.

class RunningCentralMomentsState: RunningCentralMomentsState(mean_state, sum_exponentiated_residuals)

class RunningCovariance: Holds metadata for and facilitates covariance computation.

class RunningCovarianceState: RunningCovarianceState(num_samples, mean, sum_squared_residuals)

class RunningMean: Holds metadata for and computes a running mean.

class RunningMeanState: RunningMeanState(num_samples, mean)

class RunningPotentialScaleReduction: Holds metadata for and computes a running R-hat diagnostic statistic.

class RunningPotentialScaleReductionState: RunningPotentialScaleReductionState(chain_var,)

class RunningVariance: Holds metadata for and facilitates variance computation.