sphereC

Okay thanks for the swift reply. Is it possible to keep the cache within R memory (I mostly want to use drake to manage the complexity of all the steps,...

Thanks that worked. I'm having now other issues with storr, meaning cache$archive_export (for intermediate workspace saves) gives me Error: hash 'db2ded51d81a4403d8a4bd25fa1e57ee' not found But I guess that's more something for...

Okay thanks. I was able to use the import/export method instead and that worked but now drake doesn't recognize the history anymore (unclear if I should post this in the...

My understanding is that drake does not keep any of its cache in memory in case it is a filesystem storr. So by using a file-system-storr I would arrive at...

BTW I'm seeing some strange behavior when running the in-memory cache and then running one other step (even just the repetition of same steps). It seems to reload everything (or...

Thanks! Understood and agreed on the R6 point. It was more a convenience for me, rather than a design choice I can't live without. For the lists I'll go with...

I'm using the one from github. `install_github("ropensci/drake")` It honestly seems like it's again doing some loading of memory (meaning I again see that pyramid-like loading into memory). I'll try to...

One follow-up point: When using `make(plan, targets=c('target'),memory_strategy='none',garbage_collection=TRUE)` it sometimes wants to rerun very early points, even though I clearly specify what I want to run. Is there some way to...

Okay will do. Thanks again for the quick replies!