invacost v1.1-6 (Release date: 10 February 2024) ============== * Removed deprecated functions from the package invacost v1.1-5 (Release date: 14 April 2023) ============== * Updated citation format invacost v1.1-4 (Release date: 5 February 2022) ============== * Update to invacost v4.1 invacost v1.1-3 (Release date: 10 December 2021) ============== * Added automatic checks for correcting errors in the database * Added an option to include RMSE in model graphs invacost v1.1 (Release date: 10 June 2021) ============== * Fixed a small taxonomic error in the invacost database invacost v1.1 (Release date: 8 June 2021) ============== * Updated to invacost v4.0 * Minor changes to help files and tutorial invacost v1.0 (Release date: 3 December 2020) ============== * Official release of invacost to be uploaded to the CRAN * Several minor fixes invacost v0.3-9 (Release date: 2 December 2020) ============== * Updated all help files * Changed default values in summarizeCosts() (ex-calculateRawAvgCosts()) invacost v0.3-8 (Release date: 1 December 2020) ============== * Added function getInvaCostVersion() to download older releases of InvaCost invacost v0.3-7 (Release date: 30 November 2020) ============== * Changed function names * Updated to invacost v3.0 * Changed default values in modelCosts (ex-costTrendOverTime()) invacost v0.3-5 (Release date: 3 November 2020) ============== * Fixed all CRAN checks warnings and notes * Changes to documentation before CRAN submission * Added error checks invacost v0.3-4 (Release date: 10 August 2020) ============== * Added a function to make a "pretty" summary of model summaries invacost v0.3-3 (Release date: 3 August 2020) ============== * Fixed a bug with RMSE estimation for robust models when data did not have values for all years * Added description of columns for the invacost dataset * Updated to invacost version V2-1 * Added a column with invacost version invacost v0.3-2 (Release date: 17 July 2020) ============== * Added a feature in plot-method to allow user to select which model outputs are included - This applies to either single plot or facet plot and shares model names costTrendOverTime Fixed erroneous RMSE of GAM models invacost v0.3-1 (Release date: 7 July 2020) ============== Fixed a bug in costTrendOverTime when incomplete.years.weight was set to NULL * General improvements for calculrateRawCosts: - added annual costs in the output - improved handling of the last year with a new argument invacost v0.3 (Release date: 3 July 2020) ============== * Major changes to the models in costTrendOverTime(): - added robust regression models (thanks Andrew Kramer!) - changed error estimations for ordinary least square regressions in order to account for heteroscedasticity and temporal autocorrelation - changed GAM model family to use Gaussian Location-Scale family, in order to account for heteroscedasticity - added the summary of models directly in the output of costTrendOverTime() * Added precisions about the error bands for MARS models in the help file invacost v0.2-4 (Release date: 7 April 2020) ============== Changes: * Changed the method used to manually specify graphical parameters invacost v0.2-3 (Release date: 25 February 2020) ============== Changes: * Added the number of estimates and the number of "year values" (total number of individual values spread over all years in raw costs) invacost v0.2-2 (Release date: 27 January 2020) ============== Changes: * Major change in how objects are plotted. Now, basic functions will not make any plot, and all plots will be called with plot(object). * Greatly improved plotting of invacost objects, included new possibilities and arguments: - choice between bars or points, average or total costs for raw costs - choice of a single or multiple facets for cost trends over time invacost v0.2-1 (Release date: 22 January 2020) ============== Changes: * Fixed an issue with calculateRawAvgCosts: previously the average per period was calculated on the basis of effective years (i.e., empty years were excluded from the average). Now, the average will always be based on the entire period. There is no reason to believe that incomplete year are only those for which we have no data at all, so we should not make any tentative correction based on so-called "effective years" here. Furthermore, it implied comparing different time intervals (e.g., 1 year vs. 10 years), which is generally inadequate. From now on, the function will make a raw average per period of time regardless of data completion. invacost v0.2 (Release date: 19 December 2019) ============== Changes: * Added the calculateRawAvgCosts function invacost v0.1 (Release date: 30 November 2019) ============== Changes: * First version of the package