m2beta {preputils}

Convert logit transformed M-Values of proportional data back to original 0/1 range

Description

Despite conducting analysis of proportional data in M space, for publication figures the estimated values are commonly shown in the original space (range between 0 and 1). This function provides backscaling of the M values to original space by inverting the logit transformation done by beta2m()

Usage

m2beta(M)

Arguments

M

vector or matrix holding the original data

Details

Data are transformed according to

b=\frac{2^M}{2^M+1}

Value

A named vector/matrix with same dimensions as M and transformed values

Examples

    b = 1:99 / 100
    M = beta2m(b)
    plot(b,m2beta(M))
    print(all.equal(b, m2beta(M)))