Unpredictable / long simulation runtime

[SachaRG]

Hello,

(disclaimer: I'm new to Pluto ...)

I am running simulations in Pluto using functions that are from a julia package local to my computer. These simulations run very quickly in jupyter notebook: a single simulation will take about 0.2 s, while a batch of 50 is taking about 9.5 s. However, when I run the same simulations in Pluto, I get two different results, and I can't identify what I'm doing differently in each case (the code is the same):

1. The single simulation takes about 25 s to run initially, then the batch runs relatively quickly (about 100 s, compared to 9.5 s in jupyter). Future calls of the single simulation run quickly (< 1 s). If I edit the code block for the batch, however, it runs for a long time (> 1700 s). I haven't waited until it stops running.

2. The single simulation takes about 25-45 s to run initially, then the batch runs for a long time as in (1).

If I'm getting stuck at the batch in (1) or (2), I have to quit Pluto from the command line (the interrupt button doesn't always work to interrupt the slow batch). Then I restart julia, activate my environment using julia> ] activate . (for some reason I'm having to do this each time), restart Pluto and reload the notebook. Usually (1) happens but sometimes (2).

It would be really helpful to get this running close to jupyter notebook speed, since I'm trying to create a notebook with dynamically updating plots of simulations. Any help with this would be very appreciated!