Am 03.02.19 um 02:28 schrieb Adrien Dulac: > Thus, as far as I understand, to compute the conditional posterior > distribution of the weight I set, we do |np.exp(entropy - new_entropy)|. But > as the difference is big, the exponential is always zero.
The probability is only _proportional_ to this number. As I said, the posterior obtained in this way is unormalized. Hence, it does not make sense to do this for a single edge. You have to do for more than one edge, and normalize to obtain the _relative_ probability between them. Alternatively, you can do it for several weight values for the same edge, and then normalize between them. This is explained with an example in the howto: https://graph-tool.skewed.de/static/doc/demos/inference/inference.html#edge-prediction-as-binary-classification To get normalized marginal distributions for single edges is necessary to use the network reconstruction framework, but this still needs to be updated for edge covariates. -- Tiago de Paula Peixoto <[email protected]>
signature.asc
Description: OpenPGP digital signature
_______________________________________________ graph-tool mailing list [email protected] https://lists.skewed.de/mailman/listinfo/graph-tool
