naikai / sake

sake
27 stars 12 forks source link

t-SNE errors when running sake with more than 1 cpu core #78

Open psterk opened 4 years ago

psterk commented 4 years ago

If I use more than 1 CPU core, sake fails to produce t-SNE plots with the following error:

Library Rtsne loaded.
Library Rtsne loaded in cluster.

Timing stopped at: 2.894 0.368 7.259
Error in checkForRemoteErrors(val) : 
  2 nodes produced errors; first error: Operation not allowed without an active reactive context. (You tried to do something that can only be done from inside a reactive expression or observer.)

This happens on MacOS (Catalina) and on Ubuntu 18.04, both running R 3.6.1, shiny 1.4.0, Rtsne 0.15. NMF runs happily with multiple CPUs. I'm not sure whether this can be fixed easily, but I can live with just using one CPU core when I want to generate t-SNE plots.

naikai commented 4 years ago

I'll take a look and get back to you. Thanks for reporting it.

naikai commented 3 years ago

Hi @psterk ,

Sorry for the late reply. This is the message I got when I run t-SNE using 16 cores. Can you tell me which back end you are using for running parallel jobs? Thanks.


Library Rtsne loaded.
Library Rtsne loaded in cluster.

Stopping cluster

snowfall 1.84-6.1 initialized (using snow 0.4-3): parallel execution on 16 CPUs.

Library Rtsne loaded.
Library Rtsne loaded in cluster.

Stopping cluster