Lodestar.Stats 0.5.0. This page is frozen at that release. Read the current documentation for what main says now. A link to a decision or a migration page follows main, and leaves the archive.

Home › Hypothesis tests

PearsonResult

A Pearson correlation and the p-value that goes with it.

public sealed record PearsonResult(double Statistic, double PValue)

Properties — Statistic is the correlation coefficient, in [-1, 1]. PValue is the p-value on the requested tail.

Example — the coefficient and its interval come from one call, not two.

using Lodestar.Stats;

double[] sugar = [4.0, 6.5, 5.0, 9.0, 7.5, 3.0, 8.0, 6.0];
double[] sweetness = [3.1, 5.9, 4.0, 7.4, 6.8, 2.2, 6.1, 5.2];

PearsonResult result = Pearson.Test(sugar, sweetness);
(double low, double high) = result.ConfidenceInterval();

double lower = Math.Round(low, 4);    // => 0.8783
double upper = Math.Round(high, 4);   // => 0.9961

Remarks — alongside the two public properties, a PearsonResult privately carries the number of pairs it was computed over and the tail that was asked for. They are internal because a caller only reaches them through ConfidenceInterval, which needs both: the sample size sets the standard error, and the tail decides whether the interval is half-open. A second call that re-derived them would be a second chance to disagree with the first, which is the reason scipy carries the same two on its own PearsonRResult.

Being a record, two results with the same statistic and p-value are equal — the internal fields take part in that equality too, so two results from samples of different sizes are not equal even when their two numbers coincide.

Applies to — net10.0, netstandard2.0.

See also — Pearson.Test, TestResult, TTestResult, the Python equivalence table.

Members

Member	What it does
PearsonResult.ConfidenceInterval	The confidence interval for the correlation, through the Fisher z transform.