PCRaster and Python version 2 have been loyal friends for a long time, but even good things won’t last forever. As the official support for Python 2 ends in less than three months, we no longer make efforts in building new versions of PCRaster against Python 2. The upcoming PCRaster 4.3 release (expected later this year) will exclusively support Python 3 on all supported platforms. Modellers really should migrate their model code to Python 3 in case they did not until now.
We will be in Vienna next week, our contributions are:
The LUE scientific data base for storing heterogeneous earth science data
Tue, 09 Apr, 14:00–15:45 Hall X1
Sustainable scientific software: experiences of the PCRaster research and development team
Wed, 10 Apr, 08:30–10:15 Hall X1
We are glad to announce the release of PCRaster 4.2.1. We fixed a memory leak in
pcr2numpy, thus long lasting model runs can be fired up again!
We are glad to announce the final release of PCRaster 4.2.0! We added support for Python 3, the new ‘multicore’ module providing multi-threaded local and focal operations, fixed several software issues and merged all our manuals into one documentation. For more information please visit the download page. Enjoy!
We received some user feedback since we shipped the PCRaster 4.2rc1, amongst others reporting one relevant issue in the Python module. The conversion of PCRaster non-spatial values to NumPy was defective in previous releases, this should be resolved now.
Please check out the PCRaster 4.2rc2 for Windows. In case no further issues are found the final version of PCRaster 4.2 will be released next week.
We are finalising the upcoming PCRaster 4.2 release. Briefly, the main new features are the new multicore module for PCRaster Python, support for Python 3, and enabling large file support on Windows systems.
PCRaster support for Python 3
The lifetime of Python 2 is coming to an end, Python 3 is now widely used and even conservative Linux distributions like RedHat will switch to Python 3 as default version in their upcoming major release. We now support Python 2 or 3 on Linux distributions. Python version 3.6 is now our supported version by default on Windows.
The PCRaster multicore module
The new PCRaster Python multicore module contains alternative implementations of a subset of the PCRaster operations that are capable of distributing their workload across more than one CPU core, therefore improving the runtime performance of those operations. Modifications to existing model scripts are not required. The multicore module can be enabled just by setting an environment variable.
Support for files larger than 2 GB on Windows systems
With previous PCRaster versions it was on Windows systems impossible to generate raster maps larger than 2 GB, this issue is resolved now. We also updated the PCRaster driver in GDAL, version 2.2.4 or later is required to use the new functionality.
Further improvements and Windows package
We unified the documentation of all our projects. We also spent significant efforts in improving our code base as well as the build system for different operating systems. Supporting several platforms should be more straightforward now.
Please check out the release candidate for Windows 64-bit. In case no major bugs are found, we plan to release the final version of PCRaster 4.2 within a few weeks.
Build your own PCRaster
We already mentioned earlier that we will no longer distribute compiled versions of PCRaster for Linux systems. Users will need to build their own package, but we strive to make that process simple. Debian testing users, for example, can try this:
$ sudo apt install cmake gcc g++ git libboost-all-dev libgdal-dev libncurses5-dev libpython-dev libqwt-qt5-dev libxerces-c-dev libxml2 libxml2-utils libxslt1-dev python-numpy qtbase5-dev xsdcxx python-docopt
$ mkdir pcraster42rc1 && cd pcraster42rc1
$ git clone --recursive https://github.com/pcraster/pcraster.git
$ mkdir build
$ cd build && cmake -DFERN_BUILD_ALGORITHM:BOOL=TRUE ../pcraster && cmake --build .
We will provide build instructions for other Linux distributions in another post.
We have a vacancy for a PhD student working at the intersection of computer sciences and geosciences, with important contributions to the next generation modelling software.
Information about the position can be found at:
Search for (select ‘Academic’ and ‘Faculty of Geosciences’):
PhD Parallel algorithm design for distributed geoscience applications
The vacancy is no longer available
RIVM is offering a traineeship to construct PCRaster models for the ‘Atlas Natuurlijk Kapitaal’. Information is available at (in Dutch) http://karssenberg.geo.uu.nl/traineeshipRIVM.html.