Univariate Class

This is the central class for the analysis of univariate samples of independent, identically distributed values.

To compute moments of the sample data, you can use methods such as Mean, Variance, RawMoment, and CentralMoment. Note that these are moments of the sample data, not estimates of the moments of the underlying population from which the sample was drawn.

To obtain estimates of the moments of the underlying population from which the sample was drawn, you can use methods such as PopulationMean, PopulationStandardDeviation, PopulationRawMoment, and PopulationCentralMoment. These estimates all come with associated error bars, so they return UncertainValue structures.

You can fit a sample to any number of distributions using methods such as FitToExponential(IReadOnlyListDouble), FitToLognormal(IReadOnlyListDouble), FitToNormal(IReadOnlyListDouble), FitToWeibull(IReadOnlyListDouble).

You can perform statistical tests on a single sample, such as the StudentTTest(IReadOnlyCollectionDouble, Double) or SignTest(IReadOnlyCollectionDouble, Double) to compare a sample to a reference value, or the ShapiroFranciaTest(IReadOnlyListDouble) to test a sample for normality. You can also perform statistical tests comparing multiple samples, such as the two-sample StudentTTest(IReadOnlyCollectionDouble, IReadOnlyCollectionDouble), and MannWhitneyTest(IReadOnlyListDouble, IReadOnlyListDouble), or the multi-sample OneWayAnovaTest(IReadOnlyCollectionDouble) and KruskalWallisTest(IReadOnlyListDouble).

Most of the methods in this class are extension methods that accept as a sample any type that implements the appropriate collection interface. So, for example, given sample values in a List<string> named s, you could estimate the variance of the population from which it was drawn either by s.PopulationVariance() or Univariate.PopulationVariance(s).

Reference