NEWS file for package multiblock Functions and datasets to support Smilde, Næs and Liland (2022, ISBN: 978-1-119-60096-1) "Multiblock Data Fusion in Statistics and Machine Learning - Applications in the Natural and Life Sciences". This implements and imports a large collection of methods for multiblock data analysis with common interfaces, result- and plotting functions, several real data sets and six vignettes covering a range different applications. Kristian Hovde Liland ##### Version 0.8.8.2 Date: 15 September 2024 >> Corrected errors in two-level ASCA modelling affecting plotting >> Added R2 and MSEP for MB-PLS. ##### Version 0.8.8.1 Date: 11 March 2024 >> Corrected errors in SO-PLS's cross-validation (segment use) and faulty summary when only one X-block is supplied. ##### Version 0.8.8 Date: 26 January 2024 >> Corrected error in POPLS leading to faulty explained variances and block scores. ##### Version 0.8.7 Date: 17 November 2023 >> Added block.preprocess to perform block-wise preprocessing of data. >> Added _PACKAGE to conform to changed roxygen2. ##### Version 0.8.6 Date: 04 June 2023 >> Moved several packages to Suggests and added conditional loading. >> Added option for user-defined confidences in ASCA ellipsoids. >> Corrected error in predict.rosa() (wrong intercept when using coefs argument). >> Extended documentation for coef.rosa() and predict.rosa(). >> Added support for data without dimnames in ROSA and SO-PLS. ##### Version 0.8.5 Date: 6 March 2023 >> Corrected weighting in block loop for MB-PLS. ##### Version 0.8.4 Date: 20 February 2023 >> Corrected GCA (gca) outputs by removing blockLoadings, adding blockDecomp and exchanging blockCoef. >> Added support for setting number of components per block in GCA. >> Fixed bug in MB-PLS (mbpls) when blockscale="none" (thanks to GitHub user flomehl). >> Corrected scaling of response in SO-PLS-PM (sopls_pm). ##### Version 0.8.3 Date: 29 October 2022 >> Added options for various block weights in MB-PLS and sMB-PLS. >> Added correlation loading plots (corrplot). >> Improved feedback for score- and loadingplot for SO-PLS. >> Added possibility of plotting scores from multiple blocks for SO-PLS. >> Changed computations of explained variances for DISCO. >> Fixed wrong fitted values for SO-PLS model with 1 component in first block. ##### Version 0.8.2 Date: 06 June 2022 >> Changed default behaviour of sopls_pm to simultaneously orthogonalise on intermediate blocks instead of orthogonalising sequentially. Legacy behaviour (used in some papers) can be obtained by setting simultaneous=FALSE. >> Changed default interface of MB-PLS, sMB-PLS and MB-RDA to use formula, but retained backward compatibility with new arguments X and Y. >> Changed internal handling of X-list in SO-PLS/ROSA to improve naming (excluding block names in variables). ##### Version 0.8.1 Date: 28 December 2021 >> Added NEWS to package. >> Improved documentation of PCA-GCA. >> Changed ROSA default argument internal.validation=FALSE. ##### Version 0.8.0 Date: 07 September 2021 >> Initial CRAN release of package multiblock.