Be sure that all the sample calls get the same random seed in syn_test_data

[Melkiades]

there are three instances of sample in the syn_test_data that generate the data for testing for adlb (here

chevron/R/utils.R

Line 173 in 9f3d0ce

sample(x = c("LAST", "REPLICATED", "SINGLE"), size = n(), replace = TRUE, prob = c(0.3, 0.6, 0.1)),

and below.) Please set seed of find another way to generate data. Cheers

Originally posted by @BFalquet in #418 (comment)

This is generally to avoid further unwanted updates of testing snapshots. @clarkliming @BFalquet

[edelarua]

@Melkiades as discussed with Benoit, I will add the fix for this to #418.

[BFalquet]

btw, doesn't need to be the same seed everywhere, just need to be constant (i.e. explicitly declared)

[clarkliming]

isn't this sufficient ?

chevron/R/utils.R

Line 144 in 9f3d0ce

set.seed(321)

[BFalquet]

apparently not (that was also a surprise for me). See: https://github.com/insightsengineering/chevron/pull/418/files#r1124228431

[BFalquet]

We should store syn_data as a fixed dataset, that would avoid all of this issue

[edelarua]

Closed by 3e239d1