We have been working very hard to update our line of Docker images for many groups of Renku users. In particular, some external dependencies on R/RStudio have been resolved and we are excited to finally upgrade these to recent images.
Python
Our default image uses Python 3.9 but sometimes older versions of Python are needed. We now also provide images with Python 3.8.8 and 3.7.6.
R, RStudio and R-Bioconductor
We now have base Renku images for R 4.1.0, 4.1.1, 4.1.2, devel and R-Biodoncutor 3.13, 3.14, devel which all run RStudio 2021.09.2-382. This should be a huge upgrade for those who have been waiting for it and we thank you for your patience!
Julia
Julia has been bumped to 1.7.1.
CUDA, CUDA-TF
We have base images for nearly all recent 11.X versions: 11.0, 11.1, 11.2, 11.3, 11.4, 11.5. For those using Tensorflow, we also have 2.4 with CUDA 11.0 and Tensorflow 2.7 with CUDA 11.2.
MATLAB
We have released a new Renku image based on the Renku desktop which includes MATLAB R2021b. You will need a valid MATLAB licence.
QGIS
We have also released an image with QGIS 3.18 built-in which also uses the Renku desktop.
How to get your hands on the new images
The base Renku templates have already been updated, so you can upgrade the template in Renku if you are prompted. If you would like to manually update, you can do it by changing the relevant Dockerfile(s). See Release Version 0.11.0 · SwissDataScienceCenter/renkulab-docker · GitHub for the tag names. The Desktop, MATLAB and QGIS templates have been added to the contributed images too so you can immediately start a new project with these templates!
Let us know if you have any issues or further requests.