cvmdisc (version 0.1.0)

cvmPval: cvmPval

Description

Calculate P values of CVM statistics using their asymptotic distributions

Usage

cvmPval(statistic, Msig, imhof = TRUE)

Arguments

statistic

test statistic

Msig

Matrix used to produce eigenvalues to estimate the asymptotic distributions of the test statistic

imhof

Logical. Set to be FALSE if Imhof's method is NOT used to approximate the null. The package"CompQuadForm" is required if "imhof=TRUE"

Value

P-value for test statistic

Details

cvmPval is used by groupFit to calculate test statistics for fit distributions.

See Also

groupFit: Data fitting function

Examples

# NOT RUN {
# A_squared and MsigA derived from
#(Choulakian, Lockhart and Stephens(1994) Example 3, p8)

A_squared <- 1.172932

MsigA <- matrix(c(0.05000, 0.03829, 0.02061, 0.00644,
                  0.03829, 0.30000, 0.16153, 0.05050,
                  0.02061, 0.16153, 0.30000, 0.09379,
                  0.00644, 0.05050, 0.09379, 0.30000),
                nrow = 4, ncol = 4, byrow = TRUE)

(U2Pval1 = cvmPval(A_squared, MsigA))

U_squared <- 0

MsigU <- matrix(c(0.16666667, 0.10540926, 0.0745356, 0.05270463, 0.03333333,
                  0.10540926, 0.16666667, 0.1178511, 0.08333333, 0.05270463,
                  0.07453560, 0.11785113, 0.1666667, 0.11785113, 0.07453560,
                  0.05270463, 0.08333333, 0.1178511, 0.16666667, 0.10540926,
                  0.03333333, 0.05270463, 0.0745356, 0.10540926, 0.16666667),
                nrow = 5, ncol = 5, byrow = TRUE)

(U2Pval2 = cvmPval(U_squared, MsigU, imhof = FALSE))

# }