Changes in version 1.6.1 (2026-04-18) o Bug fixes and code optimisation. Changes in version 1.6.0 (2026-01-20) NEW FEATURES o perLog: new implementation of test of differences in mean/location for compositional data based on pairwise logratios. Changes in version 1.5.0-5 (2025-06-30) MODIFICATIONS o multLN, multKM, splineKM: revised versions to remove dependence on the NADA package. o lcTest and related functions removed: to be replaced by updated procedure soon. Changes in version 1.5.0-4 (2024-06-19) MODIFICATIONS o Fix small bug in mult- functions when using labelled rows. Changes in version 1.5.0-3 (2024-03-13) MODIFICATIONS o Fix z.warning behaviour and messages. Changes in version 1.5.0-1 (2024-01-09) MODIFICATIONS o Error when all columns contain more zeros/unobs values than threshold set by z.warning controlled. Changes in version 1.5.0 MODIFICATIONS o Parametric methods requiring regular data sets (no. obs > no. vars) now warn the user about this and stop. o When no dl vector/matrix provided then column minima used automatically as thresholds and a warning message is shown. o Handling excess of zeros/unobs modified: new logical argument z.delete allows the user to decide what to do with columns/rows containing zero/unobs values over the threshold set by z.warning. Changes in version 1.4.1 (2023-08-23) MODIFICATIONS o Handling excess of zeros/unobs (z.warning argument): columns/rows individually exceeding the given zero/unobserved values proportion threshold are now automatically deleted. A warning message details the particular columns/rows involved. o Minor bugs fixed and edits to the documentation. Changes in version 1.4.0-1 (2022-03-26) NEW FEATURES o lrSVD and lrSVDplus: imputation algorithm based on singular value decomposition for censored, missing data, or both simultaneously. Particularly thought for wide data sets (no. obs < no. vars). MODIFICATIONS o Error control: dl argument checked for unique vector or matrix internally and converted if possible. o Error control: error thrown when GBM method not applicable due to only one positive value in variable. o New argument added throughout to warn about excess of zeros/unobserved values per column/row. o multRepl & others: previous parameter 'delta' is deprecated and replaced by 'frac' to avoid confusion with original notation used in multiplicative replacement method's literature. o Minor bugs fixed and edits to the documentation. Changes in version 1.3.4 (2020-03-04) MODIFICATIONS o Minor fix to meet upcoming R version requirement. o Bug found in lrDA fixed. Changes in version 1.3.3-1 (2019-12-08) MODIFICATIONS o Minor fix to meet upcoming R version requirement. Changes in version 1.3.3 (2019-11-17) NEW FEATURES o zPatterns allows to personalise display of patterns (see new bar.ordered argument). MODIFICATIONS o zPatterns: internal optimisation. o Updated documentation. Changes in version 1.3.2-2 (2019-11-13) MODIFICATIONS o lrDA: external dependence on miscF and related removed. o multKM: minor bug related to dependency on NADA smashed. Changes in version 1.3.2 (2019-06-02) NEW FEATURES o lrDA allows to impute missing data. MODIFICATIONS o Updated documentation. Changes in version 1.3.1 (2019-05-27) NEW FEATURES o multRepl allows to impute missing data (BETA). o multReplus: new function to simultaneously deal with zeros and missing data by multiplicative simple replacement (BETA). MODIFICATIONS o lrEM: multiplicative simple replacement can be also used to initiate the EM algorithm for missing data imputation. o Minor bugs fixed. o Updated documentation. Changes in version 1.2.0 (2019-02-18) NEW FEATURES o lrEMplus: new function to simultaneously deal with zeros and missing data based on the EM algorithm. MODIFICATIONS o zPatterns able to separate zero and missingness patterns. New warnings about either zeros or NAs not identified by the label argument. o Updated documentation. Changes in version 1.1.2 (2018-11-06) MODIFICATIONS o Minor bug fixed. o Updated documentation. Changes in version 1.1.1 (2017-08-09) NEW FEATURES o lrEM now allows to impute missing data. o zVarArray: variation arrays by zero/unobserved data patterns or groups. o zVarArrayError: squared relative errors in variation arrays across groups. o zVarArrayTest: homogeneity test of variation arrays across groups. o lcTest: homogeneity test of log-contrast across groups. MODIFICATIONS o Improved documentation. o Verbosity controlled throughout by suppress.print argument. o zPatterns: argument to show means (geometric or arithmetic) by pattern included. Changes in version 1.0.3-1 (2016-04-14) MODIFICATIONS o References updated and adaptation to new CRAN policies. Changes in version 1.0.3 (2014-09-12) NEW FEATURES o Support for multiple limits of detection/threshold values by component. o Choice for storing multiply imputed data sets in lrDA function. o multKM: new non-parametric imputation method added. o splineKM: visualisation of Kaplan-Meier and cubic spline smoothed empirical cumulative distribution function (helper function for multKM). o New example data set included (Water). MODIFICATIONS o Alternative imputation by multRepl in lrEM and lrDA of censoring patterns with only one observed component. o Fixed problem that caused error when a censoring pattern consisted of a single sample. o Improved documentation. Changes in version 1.0.1 (2014-06-03) o Minor bugs in error messages fixed. o cmultRepl: label argument added (default label = 0). Changes in version 1.0.0 (2014-05-20) NEW FEATURES o cmultRepl function: bayesian-multiplicative replacement of compositional count zeros included. o lrDA function: data augmentation algorithm, including multiple imputation, to replace left-censored values. o zPatterns function: summarises the patterns of unobserved values (censored, nondetects, ...) in a data set and generates a vector of labels. MODIFICATIONS o General revision and optimisation. Documentation improvements. o multLN: parameter estimates based on the normal distribution on the positive real line. Random imputation based on truncated normal instead of rejection method. o Re-scaling to preserve ratios that leaves absolute observed values unaltered when working with non-closed data now implemented for all the methods (continuous data). o The replacement methods include an argument 'label' allowing the user to enter the label (special character, number, ...) denoting unobserved value in a data set. o Error handling introduced. o New lrEM function: replaces previous alrEM function and implements both ordinary and robust EM-based imputation methods. New arguments allow to specify the method to obtain initial estimates and the convergence criterion.