minimize_nested_blockmodel_dl got an unexpected keyword argument 'multilevel_mcmc_args'
Hello, I have a question about minimize_nested_blockmodel_dl(). The below is my code: state_args = {'clabel': node_types, 'pclabel': node_types} multilevel_mcmc_args = {'niter': niter, 'beta': beta} state = gt.minimize_nested_blockmodel_dl( g, deg_corr = True, overlap = False, state_args = state_args, multilevel_mcmc_args = multilevel_mcmc_args ) And I encountered an error that "minimize_nested_blockmodel_dl got an unexpected keyword argument 'multilevel_mcmc_args'". I removed "multilevel_mcmc_args" and it works. And I also tried example "minimize_blockmodel_dl(G, multilevel_mcmc_args=dict(B_min=4, B_max=4))" in https://git.skewed.de/count0/graph-tool/-/issues/725 and I encountered the same error. I took a look of the source code of graph-tool and I did not find where is the problem. Could you please help me about it? Thanks! Best, Siwei
Am 13.07.22 um 01:28 schrieb Siwei Zhang:
Hello,
I have a question about minimize_nested_blockmodel_dl(). The below is my code:
state_args = {'clabel': node_types, 'pclabel': node_types} multilevel_mcmc_args = {'niter': niter, 'beta': beta} state = gt.minimize_nested_blockmodel_dl( g, deg_corr = True, overlap = False, state_args = state_args, multilevel_mcmc_args = multilevel_mcmc_args )
And I encountered an error that "minimize_nested_blockmodel_dl got an unexpected keyword argument 'multilevel_mcmc_args'". I removed "multilevel_mcmc_args" and it works. And I also tried example "|minimize_blockmodel_dl(G, multilevel_mcmc_args=dict(B_min=4, B_max=4))|" in https://git.skewed.de/count0/graph-tool/-/issues/725 <https://git.skewed.de/count0/graph-tool/-/issues/725> and I encountered the same error. I took a look of the source code of graph-tool and I did not find where is the problem. Could you please help me about it? Thanks!
I cannot reproduce this. What version of graph-tool are you using? -- Tiago de Paula Peixoto <tiago@skewed.de>
Sorry for the trouble, I found the issue that I was using the previous version. I built the latest version of graph-tool in my docker image and it works. And may I ask another question? In the paper, it mentioned "by making beta goes to infinity and repeated many times, which yields a reliable estimate of the maximum". May I double-check what "repeated many times" refers to? Does it refer to the number of sweeps or refer to the whole algorithm? I also noticed there is a warning of "multilevel_mcmc_sweep" in NestedBlockState: "This function performs niter sweeps at each hierarchical level once. This means that in order for the chain to equilibrate, we need to call this function several times, i.e. it is not enough to call it once with a large value of niter." I found that the high-level function "minimize_nested_blockmodel_dl" seems already done that. But I am a little bit confused, if possible, could it be more specific? Thank you so much for your help! Best, Siwei ________________________________ From: Tiago de Paula Peixoto <tiago@skewed.de> Sent: Wednesday, July 13, 2022 7:52 To: graph-tool@skewed.de <graph-tool@skewed.de> Subject: [graph-tool] Re: minimize_nested_blockmodel_dl got an unexpected keyword argument 'multilevel_mcmc_args' Am 13.07.22 um 01:28 schrieb Siwei Zhang:
Hello,
I have a question about minimize_nested_blockmodel_dl(). The below is my code:
state_args = {'clabel': node_types, 'pclabel': node_types} multilevel_mcmc_args = {'niter': niter, 'beta': beta} state = gt.minimize_nested_blockmodel_dl( g, deg_corr = True, overlap = False, state_args = state_args, multilevel_mcmc_args = multilevel_mcmc_args )
And I encountered an error that "minimize_nested_blockmodel_dl got an unexpected keyword argument 'multilevel_mcmc_args'". I removed "multilevel_mcmc_args" and it works. And I also tried example "|minimize_blockmodel_dl(G, multilevel_mcmc_args=dict(B_min=4, B_max=4))|" in https://git.skewed.de/count0/graph-tool/-/issues/725 <https://git.skewed.de/count0/graph-tool/-/issues/725> and I encountered the same error. I took a look of the source code of graph-tool and I did not find where is the problem. Could you please help me about it? Thanks!
I cannot reproduce this. What version of graph-tool are you using? -- Tiago de Paula Peixoto <tiago@skewed.de>
Am 13.07.22 um 16:43 schrieb Siwei Zhang:
And may I ask another question? In the paper, it mentioned "by making beta goes to infinity and repeated many times, which yields a reliable estimate of the maximum". May I double-check what "repeated many times" refers to? Does it refer to the number of sweeps or refer to the whole algorithm?
The whole algorithm. I also noticed there is a warning of "multilevel_mcmc_sweep"
in NestedBlockState: "This function performs niter sweeps at each hierarchical level once. This means that in order for the chain to equilibrate, we need to call this function several times, i.e. it is not enough to call it once with a large value of niter." I found that the high-level function "minimize_nested_blockmodel_dl" seems already done that. But I am a little bit confused, if possible, could it be more specific? Thank you so much for your help!
This warning is not applicable if minimize_nested_blockmodel_dl() is being used, only if you are attempting to sample from the posterior distribution (and not finding its maximum). Best, Tiago -- Tiago de Paula Peixoto <tiago@skewed.de>
participants (2)
-
Siwei Zhang -
Tiago de Paula Peixoto