Dmitry Serykh
Dmitry Serykh
If it is the case, why does it run fine with the cuda backend?
Alright, makes sense
If this number is known, however, can we make sure that the intra-group version is not chosen per default?
Currently, only the entry point named "main" is tuned. The threshold parameters irrelevant to main are set to 1. Hence, it becomes rather tedious to tune a program with multiple...
It says so in the official [documentation](https://futhark.readthedocs.io/en/latest/man/futhark-autotune.html). So it is more of a feature than a bug. You can try this [gist](https://gist.github.com/mortvest/28505ca1cba398226fb5b829c7c71a3c).
> I see. The documentation is wrong, you just need to specify the other entrypoints as well. Just replace the top of the file with the following: > > ```futhark...
Hmm, interesting. I have to look into this. Could be related to #37
You're welcome! Yes, we are currently working on the integration of the BFAST algorithm (bfast() R function). We have an early prototype of the Python backend, and it would be...
Sadly, I expect the native Python version to be a bit slower, since the most computationally intensive operations in the R version are written in C++. Multiprocessing can be utilized...
Hmm, interesting. Thank you for your input @mirt001