Changes in version 1.13.1 - Use read.csv rather than read_csv in vignettes, as read_csv was causing a hard-to-reproduce error when building vignettes. Changes in version 1.11.1 - Don't import colSums, rowSums, which have disappeared from BiocGenerics. Changes in version 1.7.1 - counts_shifts and counts_proportions now have a "typecast" argument, allowing use of memory-efficient matrix types. Changes in version 1.5.1 - weitrix_sd_confects now has an option to drop the assumption of normally distributed weighted residuals. Changes in version 1.1.10 - Fix bug with weitrix_confects due to [[ ]] <- NULL deleting elements from a list instead of storing NULL in the list. Changes in version 1.1.9 - Peaks were sometimes in reverse order in APA example, data file updated. - Try to get rid of an odd new build error about stack usage by using serial processing in vignettes. Changes in version 1.1.8 - Use geom_bin2d in weitrix_calplot scatterplots. Changes in version 1.1.7 - Add mu_min, mu_max arguments to weitrix_calibrate_all. Changes in version 1.1.6 - Use glm2, which is less prone to optimization failure. - Auto-disable parallel processing if X11 device is open. - Add SLAM-Seq vignette. Changes in version 1.1.5 - well_knotted_spline for natural splines with good choice of knots. Changes in version 1.1.4 - weitrix_confects for differential testing. - weitrix_rms_confects now called weitrix_sd_confects. - weitrix_calplot now shows mean trend and mean +/- standard deviation trend. Changes in version 1.1.3 - weitrix_calibrate_all now includes a simple scaling factor to account for residuals from a fitted model being smaller than residuals from the true model. - weitrix_calplot blue guidelines are similarly adjusted. - Add weitrix_rms_confects to find rows with confidently excessive variation. Changes in version 1.1.2 - Vignettes use weitrix_calibrate_all, demonstrate weitrix_calplot. - weitrix_calplot now uses sqrt(weight)*residual on y axis. Changes in version 1.1.1 - Switch calibration from linear model on log dispersions to using a gamma GLM. - Add weitrix_calibrate_all function for very flexible calibration. - Add weitrix_calplot to examine quality of weights. Changes in version 0.99.2 - Automatic package builder wants package to depend on R version 4.0. Changes in version 0.99.1 - weitrix now uses the BiocParallel default parallel processing, rather than the DelayedArray default. (DelayedArray does not use parallel processing by default as of version 0.14.0.) Changes in version 0.99.0 - Bioconductor submission. Changes in version 0.1.0 - Initial version.