trilearn_cta

CTA

Package	trilearn
Version	2.0.5
Language	Python
Docs
Paper	Olsson et al.[1]
Graph type	DG
Docker	bpimages/trilearn:2.0.5
Module	trilearn_cta

Description

This is the Christmas tree algorithm (CTA) for generating decomposable graph implemented in the PyPi package trilearn. This algortihm generates a decomposable graph iteratively one node at a time by making use of the junction tree represenantation. alpha and beta are sparsity parameters, where beta is the probabilty of creatinging a new node in each iteration isolated. alpha is the probabilty of connecting the new node in each iteration to another clique in a random tree traversal, given that is is not isolated (which is controlled by beta). In summary, high values of alpha, beta give denser graphs.

Example

[
  {
    "id": "trilearn_cta",
    "order": 50,
    "alpha": 0.5,
    "beta": 0.7
  }
]

[1]

Jimmy Olsson, Tatjana Pavlenko, and Felix L Rios. Sequential sampling of junction trees for decomposable graphs. Statistics and computing, 32(5):1–18, 2022. URL: https://link.springer.com/article/10.1007/s11222-022-10113-2.