reportPerf: Report performance of x-y or y-y edge detection

In topherconley/spacemap: A novel conditional graphical model for integrative genomics

Description

Assesses the performance of space and spaceMap against a known truth by reporting power and false discovery rate for either x-y OR y-y edge detection.

Usage

reportPerf(fit, truth, YY = TRUE, aszero = 1e-06, verbose = TRUE)

Arguments

fit	A numeric matrix encoding the fitted model parameter estimates.
truth	A numeric matrix encoding the true model parameters.
YY	A logical defaulting to TRUE indicating that fit and truth are symmetric matrices corresponding to estimated and true partial correlations for y-y edges. Otherwise, assume fit and truth corresponds to the estimated and true Γ regression coefficient matrix for x-y edges.
aszero	A numeric value specifying the point at which a parameter estimate should be effectively considered of zero value.
verbose	A logical value indicating whether to report the performance to the console.

Value

A numeric, named vector containing

• power or sensitivity

• fdr or false discovery rate

• mcc or Matthew's Correlation Coefficient

• tp or true positive

• fn or false negative

• fp or false positive

• tn or true negative

See Also

reportJointPerf

topherconley/spacemap documentation built on May 31, 2019, 6:36 p.m.