• Fixed S3 inheritance for the proj method of lmranks. Added it to lmranks documentation page.
  • Changed column names of coefficients in summary.lmranks from t value to z value and from Pr(<|t|) to Pr(<|z|).
  • Removed line about residual degrees of freedom from print.summary.lmranks.
  • Example in csranks now calls data(pisa2018).
  • Clarified documentation of csranks function family.
  • Improved documentation of irank, frank, etc
  • Fixed typo in an error message
  • Updated tests
  • Changed license to GPLv3
  • Added new datasets, pisa2018 and pisa2022
  • Updated CITATION and references
  • Bugfix of plot.csranks with non default indices
  • Public release with contents of 1.1 (Rank-rank linear regression)
  • Added a new mode for regression with clustered (grouped) data
  • Bugfixes, doc, new vignette
  • Optimized vcov.lmranks method
  • Added lmranks function for linear modelling of ranks using single rank covariate and possibly other, “usual” covariates
  • Implemented methods print, summary, vcov, confint, predict for lmranks output
  • Disabled a number of methods defined for ļm (like sigma, AIC or influence)
  • Release!
  • irank now raises error if NAs present and na.rm=FALSE
  • csranks now produces csranks object. List as before, but with new rank element
  • Added plot.ranking function
  • V argument (sd before 0.4) renamed to Sigma
  • Adding irank and frank functions
  • sd argument renamed to V; now accepts ONLY covariance matrix.
  • *_marg and *_simul methods removed; use * with simul option.
  • Added possibility of features to be correlated across populations. sd argument now can accept covariance matrix.
  • Added implementation of confidence sets for ranks based on multinomial data.
  • Added implementation of confidence sets for tau-best and tau-worst.
  • This is the first release of csranks.