Balázs Knakker
@bknakker.bsky.social
doing cognitive neuroscience, interested in stats and computational methods.
I use the restart hotkey in RStudio and I see the restart button in Positron, but what I really mean is that instead of just "source", there would be a "restart R + source" button - and I don't see this in Positron either. Maybe it's just me who misses this :)
January 6, 2026 at 1:59 PM
I use the restart hotkey in RStudio and I see the restart button in Positron, but what I really mean is that instead of just "source", there would be a "restart R + source" button - and I don't see this in Positron either. Maybe it's just me who misses this :)
I think, for the same reason, IDEs should have a button for "Restart R session and source script". Do you think that is a good idea? I posted this feature request for RStudio, maybe it can get some traction here: github.com/rstudio/rstu...
Restart R and source script button · Issue #15854 · rstudio/rstudio
Problem Statement When analyzing data, frequently restarting R and sourcing a script is a good idea. This can be done in 2 steps while this is a very-very frequent thing to do. Many times sourcing ...
github.com
January 6, 2026 at 12:41 PM
I think, for the same reason, IDEs should have a button for "Restart R session and source script". Do you think that is a good idea? I posted this feature request for RStudio, maybe it can get some traction here: github.com/rstudio/rstu...
You also seem to imply, but definitely 🐇 AND 💤. For me: Rr for exploring literature via citation graphs, Zz mostly for citation. I'm trying to use Zz for storing and keeping track of papers but I'm not completely satisfied. I'd put Obsidian in the stack as well, as a third leg.
June 16, 2025 at 4:37 PM
You also seem to imply, but definitely 🐇 AND 💤. For me: Rr for exploring literature via citation graphs, Zz mostly for citation. I'm trying to use Zz for storing and keeping track of papers but I'm not completely satisfied. I'd put Obsidian in the stack as well, as a third leg.
Then you probably do :) I very often need to work with quite small n.
October 12, 2023 at 9:31 PM
Then you probably do :) I very often need to work with quite small n.
No, I mean more observations (eg. subjects an/or trials).
October 12, 2023 at 9:20 PM
No, I mean more observations (eg. subjects an/or trials).
If you treat it as category, you probably need more data. As a cont. factor eg. it can be just one linear term, for cat., it's ncats-1 terms. If you add it to rfx n of parameters further explodes. This is my hunch of what could be happening, but ofc depends on the particular error you get :)
October 12, 2023 at 6:09 PM
If you treat it as category, you probably need more data. As a cont. factor eg. it can be just one linear term, for cat., it's ncats-1 terms. If you add it to rfx n of parameters further explodes. This is my hunch of what could be happening, but ofc depends on the particular error you get :)