rankr: Undertakes paired comparisons for ranking a set of...

View source: R/rankr.R

rankrR Documentation

Undertakes paired comparisons for ranking a set of coefficients, considering whether the differences between the coefficients are significant or not.

Description

The index numbers are based on the model coefficient table that comes straight out of the model, with no sorting.

The function will return a dataframe with the results for each pair of coeeficients tested.

Usage

rankr(coeftable = NULL, vcovmat = NULL, mysongidlist = NULL)

Arguments

coeftable

coefficients table from mlogit, with one row per coefficient

vcovmat

variance covariance matrix from mlogit, with one row and one column per coefficient

mysongidlist

a dataframe containing the list of song ids to be tested. It can contain other variables but only songid will be used.

Details

rankr

Examples

songstobecompared <- summary %>% slice(seq(from=1, to=92, by=10))
mycomparisons <- rankr(coeftable = results_ml_Repeatr4, vcovmat = vcovmat_ml_Repeatr4, mysongidlist = songstobecompared)
mycomparisons


alexmitrani/Repeatr documentation built on Feb. 3, 2025, 1:36 p.m.