making ocPreds more efficient

[audreyyeoCH]

Make these more efficient:

                                    user    system      elapsed
ocPredprobDist      165.489    0.012 165.510
plotOc                     57.410       0.002  57.414

cRctPredprobDist 22.048 0.006 22.055

it's probably only ocPredprobDist as the others build on it

May 2025 results :

Examples with CPU (user + system) or elapsed time > 5s
                     user system elapsed
ocPredprobDist    165.726  0.011 165.774
plotOc             58.122  0.005  58.134
ocRctPredprobDist  22.139  0.003  22.144

[danielinteractive]

Also #134 does not really close this issue, because we don't really make the simulations more efficient there, we just run less of them. So I would still keep this issue open.

[audreyyeoCH]

June results have improved on ocPredprobDist by a lot :

user system
 elapsed
plotOc            59.677  0.001  59.684
ocPredprobDist    52.850  0.003  52.858
ocRctPredprobDist 22.403  0.008  22.415

[danielinteractive]

@audreyyeoCH nice, what was the mechanism here for the speed up? more efficient code?

[audreyyeoCH]

@audreyyeoCH nice, what was the mechanism here for the speed up? more efficient code?

I am trying to figure it out myself. Those results are from plotDecision .... the files change in that PR doesn't say much.

[audreyyeoCH]

shall we try to bring the code time down further and keep this issue open ? The vignette indeed takes 1h+ though so will look at that on another existing issue.

[danielinteractive]

yeah can keep it open