if seed was passed to sample while num_chains was >1, then seed n chain i+1 would be the exact same as seed n+1 chain i-1, which means seeds n and n+1 were not independent.
init_loss was not seeded, causing slight variations in otherwise reproducible runs.
These issues are now fixed. Also changed some tests around to adjust for this fix
Fixes two issues with seeding:
seed
was passed tosample
whilenum_chains
was >1, then seedn
chaini+1
would be the exact same as seedn+1
chaini-1
, which means seedsn
andn+1
were not independent.These issues are now fixed. Also changed some tests around to adjust for this fix