‘cyt_mint_splsda’, a new function for conducting multivariateintegrative analysis using sparse partial least squares discriminantanalysis (sPLS-DA) with multiple datasets, has been added.
‘cyt_splsda’ now has new arguments:
Added new “ExampleData5” dataset which is the same as“ExampleData1” but with a batch column for demonstration of the‘batch_col’ argument in ‘cyt_mint_splsda’.
‘cyt_heatmap’ now uses pheatmap for plotting, and provides aproper legend along with additional scaling options.
Updated the vignette and README files to include examples of thenew ‘cyt_mint_splsda’ function.
Updated ‘cyt_splsda’ and ‘cyt_pca’ logic.
‘cyt_splsda’: Fixed an error occurring when only one variable hasVIP > 1 leading to not being able to conduct sPLS-DA analysis. Now,the function checks whether the number of VIP > 1 variables is below2, if that is true the sPLS-DA model of VIP > 1 is skipped.
Fixed an issue with the ‘verbose’ argument of ‘cyt_splsda’ notworking in overall analysis due to calling the wrong object.
‘cyt_errbp’ now uses ‘ggplot2’ with ‘facet_wrap’ to createmultiple error-bar plots where the p-value and effect size labels arebased on t-test comparisons between the groups.
‘cyt_errbp’, ‘cyt_bp’, ‘cyt_bp2’, and ‘cyt_skku’ received a minortheme update keeping the background white in scenarios where it mayappear transparent due to using ‘theme_minimal()’.
Fixed some grammatical and spelling issues in the READMEfile.
Vignette for the package now show plots generated that aren’tsaved to PDF files. A temporary directory is not createdanymore.
‘cyt_rf’, ‘cyt_splsda’, ‘cyt_ttest’, ‘cyt_dualflashplot’, and‘cyt_volc’ now have a logical ‘verbose’ argument to print the output ifuser wants to. This is to ensure printed output in console can be easilysuppressed.
‘cyt_xgb’ has a new logical argument called ‘print_results’ thatworks the same as the ‘verbose’ argument in other functions mentionedabove.
Reverted the previous ‘verbose’ conditionals on plots from thefunctions mentioned above as they are essential to the output. Forexample, after creating a ggplot2 object of the plots, I have left‘print(a)’ without a ‘verbose’ conditional statement.
Have removed the arguments for changing graphical parameters in‘cyt_bp’, and ‘cyt_bp2’ so it no longer requires changes to the user’sgraphical parameters.
Added a new ‘format_output’ argument to ‘cyt_anova’ and‘cyt_ttest’ to format the output as a data frame which can be printed toshow neat format, however still dependent on whether ‘verbose’ equalsTRUE or FALSE.
‘cyt_rf’, ‘cyt_splsda’, and ‘cyt_xgb’ now has a ‘seed’ argumentto set the seed for reproducibility.
Added references in description field of the DESCRIPTION file tothe multivariate methods mentioned in the format ‘Authors (Year)<doi:…>’.
Updated examples from graphical parameters changing withoutresetting properly to now the graphical parameters reverting to originalusing ‘oldpar <- par(no.readonly = TRUE)’ in the beginning andpar(oldpar) after execution of code.
Updated ‘getting_started.Rmd’ vignette where the files createdand saved are now saved to a temporary directory using ‘tempdir()’ toavoid creating files in the user’s working directory. Additionally, thevignette now uses ‘oldpar <- par(no.readonly = TRUE)’ in thebeginning and par(oldpar) after code execution to revert to originalgraphical parameters for the ‘cyt_errbp’ examples shown.
Fixed some general CRAN automated test issues.
Updated functions and examples to avoid generating PDF and PNGfiles during checks.
Functions now also generated plots within RStudio graphics deviceand provides the option to save the plots as PDF or PNG files dependingon the function.
Since this is our initial submission, our NEWS.md contains a singleentry indicating the initial release. We plan to provide more detailedchange logs in subsequent versions.
This version is the development version of CytoProfile. We arepreparing for the initial CRAN submission. Below are the features addedin this version: