Package gov.sandia.cognition.statistics.method

Provides algorithms for evaluating statistical data and conducting statistical inference, particularly frequentist methods.

See:
          Description

Interface Summary
Binner<ValueType,BinnedType> Defines the functionality for a class that assigns values to some sort of bin.
BlockExperimentComparison<DataType> Implements a null-hypothesis multiple-comparison test from a block-design experiment.
ConfidenceIntervalEvaluator<DataType> Computes a confidence interval for a given dataset and confidence (power) level
ConfidenceStatistic An interface that describes the result of a statistical confidence test.
MultipleHypothesisComparison<TreatmentData> Describes the functionality of an algorithm for accepting or rejecting multiple null hypothesis at the same time.
MultipleHypothesisComparison.Statistic Statistic associated with the multiple hypothesis comparison
NullHypothesisEvaluator<DataType> Evaluates the probability that the null-hypothesis is correct.
 

Class Summary
AbstractConfidenceStatistic Abstract implementation of ConfidenceStatistic.
AbstractMultipleHypothesisComparison<TreatmentData,StatisticType extends MultipleHypothesisComparison.Statistic> Partial implementation of MultipleHypothesisComparison
AbstractMultipleHypothesisComparison.Statistic Partial implementation of MultipleHypothesisComparison.Statistic
AbstractPairwiseMultipleHypothesisComparison<StatisticType extends AbstractPairwiseMultipleHypothesisComparison.Statistic> A multiple-hypothesis comparison algorithm based on making multiple pair-wise null-hypothesis comparisons.
AbstractPairwiseMultipleHypothesisComparison.Statistic Result from a pairwise multiple-comparison statistic.
AdjustedPValueStatistic A multiple-comparison statistic derived from a single adjusted p-value.
AnalysisOfVarianceOneWay Analysis of Variance single-factor null-hypothesis testing procedure, usually called "1-way ANOVA".
AnalysisOfVarianceOneWay.Statistic Returns the confidence statistic for an ANOVA test
BernoulliConfidence Computes the Bernoulli confidence interval.
BonferroniCorrection The Bonferroni correction takes a pair-wise null-hypothesis test and generalizes it to multiple comparisons by adjusting the requisite p-value to find significance as alpha / NumComparisons.
ChebyshevInequality Computes the Chebyshev Inequality for the given level of confidence.
ChiSquareConfidence This is the chi-square goodness-of-fit test.
ChiSquareConfidence.Statistic Confidence Statistic for a chi-square test
ConfidenceInterval Contains a specification for a confidence interval, that is, the solution of Pr{ lowerBound <= x(centralValue) <= upperBound } >= confidence
ConvexReceiverOperatingCharacteristic Computes the convex hull of the Receiver Operating Characteristic (ROC), which a mathematician might call a "concave down" function.
DistributionParameterEstimator<DataType,DistributionType extends ClosedFormDistribution<? extends DataType>> A method of estimating the parameters of a distribution using an arbitrary CostFunction and FunctionMinimizer algorithm.
FieldConfidenceInterval This class has methods that automatically compute confidence intervals for Double/double Fields in dataclasses.
FisherSignConfidence This is an implementation of the Fisher Sign Test, which is a robust nonparameteric test to determine if two groups have a different mean.
FisherSignConfidence.Statistic Contains the parameters from the Sign Test null-hypothesis evaluation
FriedmanConfidence The Friedman test determines if the rankings associated with various treatments are equal.
FriedmanConfidence.Statistic Confidence statistic associated with the Friedman test using the tighter F-statistic.
GaussianConfidence This test is sometimes called the "Z test" Defines a range of values that the statistic can take, as well as the confidence that the statistic is between the lower and upper bounds.
GaussianConfidence.Statistic Confidence statistics for a Gaussian distribution
HolmCorrection The Holm correction is a uniformly tighter bound than the Bonferroni/Sidak correction by first sorting the pair-wide p-values and then adjusting the p-values by the number of remaining hypotheses.
HolmCorrection.Statistic Test statistic from the Shaffer static multiple-comparison test
ImportanceSampling Importance sampling is a technique for estimating properties of a target distribution, while only having samples generated from an "importance" distribution rather than the target distribution.
InverseTransformSampling Inverse transform sampling is a method by which one can sample from an arbitrary distribution using only a uniform random-number generator and the ability to empirically invert the CDF.
KolmogorovSmirnovConfidence Performs a Kolmogorov-Smirnov Confidence Test.
KolmogorovSmirnovConfidence.Statistic Computes the ConfidenceStatistic associated with a K-S test
MannWhitneyUConfidence Performs a Mann-Whitney U-test on the given data (usually simply called a "U-test", sometimes called a Wilcoxon-Mann-Whitney U-test, or Wilcoxon rank-sum test).
MannWhitneyUConfidence.Statistic Statistics from the Mann-Whitney U-test
MarkovInequality Implementation of the Markov Inequality hypothesis test.
MaximumLikelihoodDistributionEstimator<DataType> Estimates the most-likely distribution, and corresponding parameters, of that generated the given data from a pre-determined collection of candidate parameteric distributions.
MaximumLikelihoodDistributionEstimator.DistributionEstimationTask<DataType> Estimates the optimal parameters of a single distribution
MultipleComparisonExperiment A multiple comparisons experiment that does a block comparison and then a post-hoc test.
MultipleComparisonExperiment.Statistic Result of running the MultipleHypothesisComparison hypothesis test
NemenyiConfidence The Nemenyi test is the rank-based analogue of the Tukey multiple-comparison test.
NemenyiConfidence.Statistic Statistic from Nemenyi's multiple comparison test
ReceiverOperatingCharacteristic Class that describes a Receiver Operating Characteristic (usually called an "ROC Curve").
ReceiverOperatingCharacteristic.DataPoint Contains information about a datapoint on an ROC curve
ReceiverOperatingCharacteristic.DataPoint.Sorter Sorts DataPoints in ascending order according to their falsePositiveRate (x-axis)
ReceiverOperatingCharacteristic.Statistic Contains useful statistics derived from a ROC curve
ShafferStaticCorrection The Shaffer Static Correction uses logical relationships to tighten up the Bonferroni/Sidak corrections when performing pairwise multiple hypothesis comparisons.
ShafferStaticCorrection.Statistic Test statistic from the Shaffer static multiple-comparison test
SidakCorrection The Sidak correction takes a pair-wise null-hypothesis test and generalizes it to multiple comparisons by adjusting the requisite p-value to find significance as alpha / NumComparisons.
StudentTConfidence This class implements Student's t-tests for different uses.
StudentTConfidence.Statistic Confidence statistics for a Student-t test
StudentTConfidence.Summary An implementation of the Summarizer interface for creating a ConfidenceInterval
TreeSetBinner<ValueType extends Comparable<? super ValueType>> Implements a Binner that employs a TreeSet to define the boundaries of a contiguous set of bins.
TukeyKramerConfidence Tukey-Kramer test is the multiple-comparison generalization of the unpaired Student's t-test when conducting multiple comparisons.
TukeyKramerConfidence.Statistic Statistic from Tukey-Kramer's multiple comparison test
WilcoxonSignedRankConfidence This is a Wilcoxon Signed-Rank Sum test, which performs a pair-wise test to determine if two datasets are different.
WilcoxonSignedRankConfidence.Statistic ConfidenceStatistics associated with a Wilcoxon test
 

Annotation Types Summary
ConfidenceTestAssumptions Describes the assumptions and other information of a statistical confidence test.
 

Package gov.sandia.cognition.statistics.method Description

Provides algorithms for evaluating statistical data and conducting statistical inference, particularly frequentist methods.

Since:
3.0
Author:
Kevin R. Dixon