Generating Bayesian network graph with dsc file I'm using a free version of Bayesian network software called Netica. It allows only 15 nodes for the free version. Do you know any other software or R package that generate a kind of graph below using a dsc file?

I know R package called bnlearn has a function read.dsc for dsc files, but am not sure how to make a graph. If you know how you can get R to make this kind of or similar graph (with probabilities of influence in each node box), please walk me through R scripts. 
 A: You could try SamIAm. It's a great free program I've been using constantly since I first downloaded it a year ago. I'm not sure if it has any node limits, but I've had Bayesian networks with 50+ nodes, and it's handled them fine. It's really a nice program; I've been able to do almost all of my calculations and work within the application. 
A: I address your question with respect to the R environment, as you originally asked, rather than using third-party software.
The R package DiagrammeR could be a good solution. It would allow you to represent the network as a DAG, with nodes, arcs, and node-title. In terms of being able to represent the the beliefs for each node-state, I see two options.


*

*From what I can see in the DiagrammeR documentation, you are also able to represent data-frames as individual nodes. So if you were able to generate tables of node-state belief-levels, you could represent the belief levels as tables. See the package vignette for details on how to do represent R objects within diagrams.

*You could plot the belief-bars for each node, then represent those plots in place of the relevant node.
Here are some links for DiagrammeR:


*

*http://rich-iannone.github.io/DiagrammeR/

*https://cran.r-project.org/web/packages/DiagrammeR/index.html
Good luck!
A: Weka doesn't have a node limit, and in their BayesNet editor you can do something like that, but it's a little clumsy. SamIAm  can read a .dsc file (Windows version of SamIAm) or a Weka bif xml file (Mac version of SamIAm) and save to other formats like Huguin .net, which R's bnlearn library can read.
