To switch the inference algorithm to EPIS use set_bayesian_algorithm with pysmile.BayesianAlgorithmType.EPIS_SAMPLING:
net.set_bayesian_algorithm(pysmile.BayesianAlgorithmType.EPIS_SAMPLING)
Regarding explicit relevance targets in the network, here's the info from the wrappers manual
Network ...
Search found 1473 matches
- Thu Jul 03, 2025 1:10 pm
- Forum: SMILE
- Topic: Network size limitation of SMILE package
- Replies: 3
- Views: 63
- Wed Jul 02, 2025 10:54 am
- Forum: SMILE
- Topic: Does SMILE support hybrid/mixed network (with both discrete and continuous variables) parameter learning/estimation?
- Replies: 3
- Views: 213
Re: Does SMILE support hybrid/mixed network (with both discrete and continuous variables) parameter learning/estimation?
The inference in fully continuous and hybrid networks works the same way. If there's no evidence for nodes with parents which do not have evidence, forward sampling is used (this includes the hybrid case).
If there's evidence in nodes which have parents without evidence, the network will be ...
If there's evidence in nodes which have parents without evidence, the network will be ...
- Wed Jul 02, 2025 10:51 am
- Forum: SMILE
- Topic: How to do manipulation of a variable in the network
- Replies: 3
- Views: 118
Re: How to do manipulation of a variable in the network
Network.set_controlled_value works just like set_evidence - you specify the node and the outcome. Calling this function is the equivalent of using 'Control Value' in GeNIe - in GeNIe the Control Value submenu has the entries for all node outcomes.
- Wed Jul 02, 2025 10:50 am
- Forum: SMILE
- Topic: Network size limitation of SMILE package
- Replies: 3
- Views: 63
Re: Network size limitation of SMILE package
Error -42 is caused by insufficient memory in exact inference algorithm. This algorithm creates temporary data structure (the join tree), which can become very large, depending on the connections in the network.
You can switch to approximate sampling inference (EPIS), or set explicit targets in the ...
You can switch to approximate sampling inference (EPIS), or set explicit targets in the ...
- Tue Jul 01, 2025 1:08 pm
- Forum: SMILE
- Topic: Annotation in rSMILE
- Replies: 5
- Views: 96
Re: Annotation in rSMILE
BTW, if you want to move this conversation into private space, email me or send me a forum PM.
- Fri Jun 27, 2025 10:33 pm
- Forum: SMILE
- Topic: is there a way to output goodness-of-fit metric such as BIC, log likelihood score?
- Replies: 3
- Views: 161
Re: is there a way to output goodness-of-fit metric such as BIC, log likelihood score?
One more thing, we're now working on additional structure learning algorithms and there will be a way to calculate BIC for network+dataset directly.
- Fri Jun 27, 2025 10:19 pm
- Forum: SMILE
- Topic: How to do manipulation of a variable in the network
- Replies: 3
- Views: 118
Re: How to do manipulation of a variable in the network
Consider using Network.set_controlled_value. For more info on arc cutting/value control, please refer to GeNIe manual, chapter "Controlling values". In the PDF version of the manual, the chapter number is 6.2.8.
- Wed Jun 25, 2025 3:42 pm
- Forum: SMILE
- Topic: is there a way to output goodness-of-fit metric such as BIC, log likelihood score?
- Replies: 3
- Views: 161
Re: is there a way to output goodness-of-fit metric such as BIC, log likelihood score?
BTW, you can use GeNIe to calculate loglikelihood for a dataset/network without learning. This functionality is not exposed in PySMILE, however.
- Wed Jun 25, 2025 3:34 pm
- Forum: SMILE
- Topic: is there a way to output goodness-of-fit metric such as BIC, log likelihood score?
- Replies: 3
- Views: 161
Re: is there a way to output goodness-of-fit metric such as BIC, log likelihood score?
You can obtain the log likelihood after EM by calling the EM.get_last_score() method.
BayesianSearch.get_last_score() returns the Bayesian score after learning.
Currently there's no method to calculate BIC from the dataset and network without learning.
BayesianSearch.get_last_score() returns the Bayesian score after learning.
Currently there's no method to calculate BIC from the dataset and network without learning.
- Fri Jun 20, 2025 1:05 pm
- Forum: SMILE
- Topic: Does SMILE support hybrid/mixed network (with both discrete and continuous variables) parameter learning/estimation?
- Replies: 3
- Views: 213
Re: Does SMILE support hybrid/mixed network (with both discrete and continuous variables) parameter learning/estimation?
I confirm that SMILE does not support parameter learning in hybrid networks.
- Wed Jun 18, 2025 10:39 am
- Forum: SMILE
- Topic: error message regarding /lib64/libm.so.6
- Replies: 5
- Views: 928
Re: error message regarding /lib64/libm.so.6
Correct, the .so file from the .tar.gz archive is for the scenario where you do not use pip.
- Tue Jun 17, 2025 2:02 pm
- Forum: SMILE
- Topic: error message regarding /lib64/libm.so.6
- Replies: 5
- Views: 928
Re: error message regarding /lib64/libm.so.6
I have sent you a download link using private forum message.
- Mon Jun 16, 2025 12:13 pm
- Forum: SMILE
- Topic: how default probability number in the network with only categorical variable is set?
- Replies: 3
- Views: 1096
Re: how default probability number in the network with only categorical variable is set?
Later we call the EM algorithm for parameter learning on this constructed network `cate_net`, the initial numbers like 0.5, 0.5, 0, 0 etc does not matter right?
If you have the complete dataset (no missing entries), the choice of initial parameters does not matter, unless you set the equivalent ...
If you have the complete dataset (no missing entries), the choice of initial parameters does not matter, unless you set the equivalent ...
- Mon Jun 16, 2025 12:09 pm
- Forum: SMILE
- Topic: My example of simulating categorical model and parameter learning
- Replies: 3
- Views: 1225
Re: My example of simulating categorical model and parameter learning
If you are attempting to learn parameters using a dataset without missing data, the selection of initial parameters (uniformize vs randomize vs keep original) does not matter - the node probabilities are calculated by counting records in the data. Just instantiate the EM object and call em.Learn ...
- Mon Jun 16, 2025 12:05 pm
- Forum: SMILE
- Topic: error message regarding /lib64/libm.so.6
- Replies: 5
- Views: 928
Re: error message regarding /lib64/libm.so.6
Which Linux distro runs on your cluster?