Closed alecksphillips closed 11 months ago
Need to check this doesn't have a performance impact, including on MPI
Edit:
I've run 100 iterations of examples/mpi/decision_tree_example.py
with -n 4 with differing seeds before and after adding progress bars, there doesn't appear to be any effect on runtime.
Without progress bars: SMC mean run-time: 5.1817866429999775 SMC median run-time: 5.091337150000072 SMC std run-time: 0.40894942106622356 SMC min run-time: 4.464644400000907 SMC max run-time: 6.2240012000002025
With progress bars: SMC mean run-time: 5.611294212999983 SMC median run-time: 5.663562350000575 SMC std run-time: 0.4113942067145501 SMC min run-time: 4.843810599999415 SMC max run-time: 6.362878300000375
Add progress bars for SMC/MCMC iterations using tqdm. Add "verbose" parameter to DiscreteVariableSMC and DiscreteVariableMCMC (default = True) which enables/disables showing the progress bar.