Am 15.01.20 um 11:15 schrieb Davide Cittaro:
> Hello everybody,
> I'm new to graph-tool and nSBM, so forgive my naive question. We are still 
> trying to understand how parameter influence our outcome. My first question is
> 
> is
> 
> state = gt.minimize_nested_blockmodel_dl(g)
> state.mcmc_sweep(niter=100)
> 
> Equal to
> 
> state = gt.minimize_nested_blockmodel_dl(g, mcmc_args=dict(niter=100))
> 
> ? I’m asking as the documentation executes the two steps, but the
> minimization function accepts parameters for MCMC sweep step.

No, these are not the same thing.

The function minimize_nested_blockmodel_dl() employs an aglomerative
heuristic which alternates between merging groups and moving nodes
between groups, and doing a bisection search for the optimal number of
groups. The mcmc_args argument controls only the moving of nodes between
groups.

The mcmc_sweep() function only performs moves of nodes between groups,
nothing else.

-- 
Tiago de Paula Peixoto <[email protected]>

Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
graph-tool mailing list
[email protected]
https://lists.skewed.de/mailman/listinfo/graph-tool

Reply via email to