Changes in version 0.2.7 (2026-06-02) - Bug fix: Fixed namespace resolution error with lifecycle::deprecated() function calls that was causing "could not find function deprecated" errors for users. All function parameters now properly use lifecycle::deprecated() instead of deprecated(). - Bug fix: train_spectra() now correctly populates RMSEcv and R2cv for rf, svmLinear, and svmRadial model methods. Previously these were always NA for non-PLS models. - Bug fix: Replaced deprecated tidyr and dplyr functions across plot_spectra(), aggregate_spectra(), train_spectra(), and the ikeogu.2017 example to resolve deprecation warnings. - Internal: Fixed trainControl() configuration in train_spectra() — changed method from "repeatedcv" (with no repeats set) to "cv" and removed a malformed seeds argument that was being silently ignored. Reproducibility is maintained via the existing set.seed() call. - Bug fix: The final RF model returned by train_spectra() is now trained with 500 trees (the randomForest default) instead of tune.length (≤5). The tuned mtry from training iterations is now correctly applied to the final model. - Bug fix: The final SVM model returned by train_spectra() no longer incorrectly uses the unique.id column as a predictor. Training data is now subset to reference and spectral columns only, consistent with iterative model training. - Bug fix: The final PLS model returned by train_spectra() is now fit using the modal best ncomp from training iterations rather than tune.length. predict_spectra() now reads ncomp directly from the model object instead of the stats file. - Bug fix: train_spectra() now correctly subsets partitioned data using column names rather than pre-computed indices, preventing an undefined columns selected crash when cv.scheme is used (the genotype column is removed by format_cv() before subsetting). - Bug fix: train_spectra() now correctly determines the training data for the final model when cv.scheme is used. It calls format_cv() one final time to obtain the training set defined by the chosen scheme: CV0/CV00 train on trial2 + trial3; CV1 trains on t1.b + t2.b; CV2 trains on t1.b + trial2. In all schemes the test set is t1.a (held-out genotypes from trial1). - Bug fix: test_spectra() now correctly forwards best.model.metric and seed to train_spectra(). Previously these arguments were accepted but silently ignored. - Bug fix: train_spectra() now calls set.seed() before createDataPartition(), ensuring that stratified train/test splits are fully reproducible. Previously, set.seed() was called after createDataPartition(), so the partition depended on whatever the random state happened to be at call time. Note on reproducibility: users who called test_spectra() with a non-default seed were previously receiving results generated with seed = 1 regardless of the value they set. To reproduce old results, set seed = 1 explicitly. This will restore previous iterative performance statistics (RMSEp, R2p, etc.), but the returned $model object will still differ due to the RF, SVM, and PLS final model fixes also introduced in this version. - Bug fix: Removed invalid ntree and mtry arguments from predict.randomForest() calls in train_individual_model() and predict_spectra(). These arguments are not accepted by predict.randomForest() and were silently ignored. - Bug fix: Fixed corrupted importance output when running test_spectra() with multiple pretreatments and an SVM model. Previously, cbind() on a NULL importance element produced a garbage 1×1 matrix instead of NULL. Changes in version 0.2.6 (2025-10-30) - Bug fix: plot_spectra() no longer returns an error when detect.outliers is set to FALSE and no alternative title is provided via the alternate.title parameter (#29). - Bug fix: Fixed compatibility issue with updated spectacles package (v0.5.5) that was causing data frame construction errors in model performance calculations. - Fixed: Temporary CRAN archive issue with the dependency spectacles resolved (#31). - The dependency spectacles is now restored on CRAN. - waves is fully compatible with the restored version. - Performance improvements: - Optimized cross-validation loops in train_spectra() using vectorized indexing and preallocated result structures. - Optimized matrix operations in train_spectra() for faster column selection and reduced memory overhead. - Added shared utility functions to eliminate code duplication between train_spectra() and test_spectra(). - When return.distances = TRUE, the h.distance column is now located between metadata and spectra in the returned data.frame (#28). - Internal code improvements: - Significantly refactored train_spectra() and test_spectra() to reduce cyclomatic complexity (#26). - train_spectra() complexity reduced from 32+ to 25 - test_spectra() complexity reduced to 11 - Extracted shared utility functions: handle_deprecations(), validate_inputs(), partition_data(), train_individual_model(), calculate_performance(), and create_cv_control(). - Improved code maintainability and reduced duplication through function decomposition. - Added robust error handling for spectacles package compatibility. Changes in version 0.2.5 (2023-12-12) - Bug fix: predict_spectra() no longer returns error when running example code (#25). Changes in version 0.2.4 (2022-03-29) - Bug fix: Different CV schemes no longer return the same results (#20). - When cv.scheme is set to "CV2" and "CV0" and there are no overlapping genotypes between "trial1" and "trial2", format_cv() now returns NULL. Previously, results would be returned even if no overlap was present, resulting in incorrect CV scheme specification. - format_cv() parameter cv.method is now the boolean parameter stratified.sampling for consistency with other waves functions. - plot_spectra() no longer requires a column named "unique.id". Changes in version 0.2.3 (2022-03-07) - Bug fix: save_model() output now works correctly with predict_spectra(). - Bug fix: train_spectra() no longer returns an error when stratified.sampling = F. - In train_spectra(), stratified random sampling of training and test sets now allows the user to provide a seed value for set.seed(). For random (non-stratified) sampling of training and test sets, seed is set to the current iteration number. - Minor documentation updates added. Changes in version 0.2.2 (2022-02-18) - Bug fix: model.method = "svmLinear and model.method = "svmRadial no longer return an error when used in train_spectra() or test_spectra(). Changes in version 0.2.1 (2022-02-03) - Bug fix: test_spectra() now returns trained model correctly when only one pretreatment is specified. - Change the gap-segment derivative pretreatment to retain compatibility with prospectr. In the upcoming version of prospectr, the gapDer function only accepts odd values for the segment argument in order to properly compute the convolution filter. - Default plot title for plot_spectra() is now NULL (no title) if detect.outliers is set to FALSE. - Column names in output list item $summary.model.performance from test_spectra() now include underscores rather than periods for easier parsing. - Update website - New vignette: vignette("waves") Changes in version 0.2.0 (2022-01-21) - Update all files to conform to the tidyverse style guide (#6). - All functions renamed to match tidystyle. Old functions names work will be dropped after this version: - AggregateSpectra -> aggregate_spectra() - DoPreprocessing -> pretreat_spectra() - FilterSpectra -> filter_spectra() - FormatCV-> format_cv() - PlotSpectra()-> plot_spectra() - SaveModel()-> save_model() - TestModelPerformance()-> test_spectra() - TrainSpectralModel()-> train_spectra() - "Preprocessing" has been renamed "Pretreatment" to minimize confusion with physical preprocessing of samples prior to scanning. Arguments have been renamed to reflect these changes (preprocessing is now pretreatment). - Added more informative error message and documentation for random forest tune length (tune.length must be set to 5 when model.algorithm == "rf"). - Additional flexibility for plot_spectra() including color and title customization and the option to forgo filtering (#5). - Named list output for all functions to enable easier access to individual elements. - Always return model and variable importance results with train_spectra() and test_spectra(). - Add variable importance for PLSR (#9). - Enable selection of k for k-fold cross-validation within the training set. Previously, k was fixed at 5 (#10). - save_model() now automatically selects the best model if provided with multiple pretreatments. - Code simplified and streamlined to facilitate future updates. - Export predicted values as well as performance statistics for each training iteration (#11). - wavelengths is no longer a required argument for any of the waves functions. - The proportion of samples to include in the training set can now be selected with the argument proportion.train. Previously, this proportion was fixed at 0.7 (#13). - Bug fix: aggregate_spectra() now allows for aggregation by a single grouping column (#14). - The parameter save.model in the function save_model() has been renamed to write.model for clarity. Changes in version 0.1.1 (2021-04-21) - Bug fix: SVM Linear and SVM Radial algorithms no longer return errors in TrainSpectralModel(). - Bug fix: Random Forest variable importance no longer returns error in TrainSpectralModel() or when preprocessing = TRUE in TestModelPerformance() (#7). - Output for random forest variable importance now includes "Pretreatment" and "Iteration" columns. - PlotSpectra() now allows for missing data in non-spectral columns of the input data frame. - waves now has an associated paper in the Plant Phenome Journal! The citation for this paper should be used if waves is used in a paper - see citation("waves") for details. Changes in version 0.1.0 (2020-09-17) Initial package release