RLumShiny Release 0.2.7

It has been a while since our previous update concerning RLumShiny. However, work on that front has continued and in the last month we had not only one but two releases (more on that below).

The work leading to the RLumShiny 0.2.7 release concentrated on adding three new apps and improving the internal infrastructure to import more file types.

The new apps are the following:

• doseresponsecurve to plot a dose-response curve for luminescence data (based on fit_DoseResponseCurve() and plot_DoseResponseCurve())

• portableOSL to analyse CW-OSL curve data produced by a SUERC portable OSL reader and produces a combined plot of OSL/IRSL signal intensities, OSL/IRSL depletion ratios and the IRSL/OSL ratio (based on analyse_portableOSL())

• sarCWOSL to a standard SAR protocol measurements introduced by Murray and Wintle (2000) with CW-OSL curves (based on calc_analyse_SAR.CWOSL())

These new apps (and a few of the existing ones, namely fastratio, finitemixture, huntley2006, lmcurve) now implement a new way of importing files, which extends the format supported beyond CSV files. This feature works in most cases, but it’s still to be considered in beta status, as more testing may be necessary to iron out some corner cases. However, this is a first step towards making the package be usable for realistic analyses.

There was also an important improvement concerning error/warning feedback to the user: while before they would only appear in the R terminal from which the app was started, now they are also shown as notification in the app itself. This allows for more direct feedback, and should help users understand better why things go wrong.

At this stage, however, these notifications are at times a bit flickery and other times too sticky. In other words, depending on user actions they may be removed too quickly (even though the problem persists) or they stay too long after the app settings have been corrected. Improvements in their behaviour will be part of future releases.

Effectively, all of the changes described above happened in RLumShiny 0.2.6. This major release contained a large number of updates, but unfortunately it also contained a fatal bug that prevented any application from actually starting. The development work for that release consolidated some code into helper functions to guarantee a consistent behaviour across all apps.

However, it turns out that shiny has a non-conventional way of dealing with internal functions. While in traditional R packages internal function can be called directly from anywhere else in the package, for packages based on shiny they must be called explicitly with the package qualifier (for example, RLumShiny:::tryNotify()). This was not noticed until version 0.2.6 was already on its way to CRAN. The hot-fix release, already ready before Christmas, was then delayed by the closure of CRAN during the Christmas holidays.

Looking forward, our main task on the RLumShiny front is to add functionality to subset the data from within the graphical interface. At this stage, we believe this to be the only remaining blocker before RLumShiny can be effectively used for exploration and analysis of real luminescence dataset.