We had two fantastic talks by Klaus Jacobi and M.eik Michalke. Klaus talked about Eliminating cloud pixels in satellite images via chronological interpolation and Meik presented his new roxyPackage package, which makes it even easier to maintain R packages with roxygen2.
Eliminating cloud pixels in satellite images via chronological interpolationKlaus gave a great case study about his consulting work as a water engineer in Asia, where he used R to analyse the snow melting process in the mountains of Pakistan.
|Images of snow melting over time. Klaus Jacobi|
About half of the fresh water in some regions of Pakistan is sourced from snow meting during the spring season. The water is captured in dams and as the snow only melts slowly the water supply is much more predictable than the water from the monsoon season, which is the other key source of fresh water during the year.
Satellite images are used to measure the surface covered with snow and to forecast the water supply. Of course snow often coincides with clouds. Hence, many of the pictures don't show a lot at all. Klaus showed us how he used R to develop a script which reads daily raster images of the regions and compares the
Klaus kindly made his presentation available via his Dropbox folder and also created an animation of the snow melting process, which you can access here.
Creating packages with roxyPackageMeik is an R user for quite some time and he is also involved in the RKward project. Meik maintains several R packages, so he knows first hand the challenges of keeping all files up-to-date and the documentation in-line with the developments.
Although roxygen2 has made the documentation of R packages much easier, there are still some tasks which need to be done more or less manually, e.g. changing dates and version numbers in DESCRIPTION, Citation and the package Rd-file.
This is the space of roxyPackage as it deals with the package building process and creation of all the necessary files. And here is the clever bit: the packages built are stored in a fully functional local R package repository which can be synced to a web server to share them with others. This includes the generation of browsable HTML pages similar to CRAN, with support for RSS feeds from the ChangeLog. Thus, it has a different purpose than the devtools package maintained by Hadley Wickham.
For more details see the package vignette of roxyPackage. Meik's presentation is also available from this web site.