local_slope: Calculate and test the local slope of the plot at large...

In ManyTests: Multiple Testing Procedures of Cox (2011) and Wong and Cox (2007)

Description

Calculates the effective slope of the plot at large values and tests the deviation of the largest value from that line (Cox, 2011).

Usage

local_slope(p, k)

Arguments

p	Vector of p-values.
k	Number of 'top' k values.

Value

local_slope	The estimated local slope of the plot at large values.
test_statistic	The value of the test statistic.
Fvalue	The upper 5% value of the F distribution with 2 and 2k degress of freedom, which is the distribution of the test statistic under the null hypothesis.
pvalue	The p-value of the test.

Author(s)

Christiana Kartsonaki

References

Cox, D. R. (2011). Discovery: a statistical perspective. Phystat Conference CERN. <doi:10.5170/CERN-2011-006>

See Also

plot_pvalues

Examples

# generate a vector of p values
p <- runif(100, 0, 1)

local_slope(p, 10)

Example output

$local_slope
[1] 1.088289

$test_statistic
[1] 0.9199858

$Fvalue
[1] 3.492828

$pvalue
[1] 0.4147432

ManyTests documentation built on May 1, 2019, 10:16 p.m.