I had two conflicting aims when designing this suite of packages: easy of use and good performance. It has not been always possible to achieve both with the same code. Because of this, the defaults tend to favour easy of use and precision. However, the alternative functions or arguments allow the tuning for performance at the expense of easy of use. I aimed at having an easy to use command-line user-interface, because I wanted to develop packages that could be used both for teaching photobiology and for real data analysis, including scripting. Performance is important in our own research as we have close to a million daylight spectra to analyse in detail, starting from raw spectrometer detector counts.
While writing the packages I learnt a lot about programming R and some of the internals of packages like data.table and ggplot2. Because of this I rewrote quite a lot of the code at least twice. Many people, knowingly or not have been helping with this project. From companies allowing the use of data sets, developers of R itself and other packages, and all those who answered to my own and other people’s questions in Stackoverflow and other web forums. Of course everybody in my research group had some influence.
In parallel with the writing of the code I have been writing a Handbook to accompany it, with Andreas Albert as co-author and other people contributing ideas and feedback. Simply using the packages in the handbook and trying to explain how to use them, made me re-think many of the details of the user-interface, I think that the basic interface is almost settled by now, but there is still much to do. In particular, although many of the code examples in the new Handbook are written, the explanatory text and the explanations about the math behind the calculations are still mostly missing.