Entering edit mode
Siebourg Juliane
▴
10
@siebourg-juliane-5379
Last seen 10.2 years ago
Hi,
In the package 'nem' I encountered a problem when trying to learn a
nem as the bayesian network version.
I think this results from different spellings of the word
'continuous'. Setting the the mode in the default parameters suggests
'continous_Baysian', which gets checkt in the function nem(). However
the internal function nem.BN checks for 'continuous_Baysian' in order
to run.
Here a toy example with the errors I get:
Phi = sampleRndNetwork(paste("S",1:5,sep=""))
D = sampleData(Phi, 100, type="density")$D
hyper=set.default.parameters(paste("S",1:5,sep=""))
hyper$mode="continous_Bayesian"
resultBN = nem(D, inference="BN.greedy", control=hyper)
# gives: Error in score_BN(BN, mode = mode, loc = rep, ...) :
inexistant mode
hyper$mode="continuous_Bayesian"
resultBN = nem(D, inference="BN.greedy", control=hyper)
# gives: Error in nem.BN(D, "greedy", mode = control$mode, Pm =
control$Pm, lambda = control$lambda, :
# Unknown mode ' continuous_Bayesian '
Maybe I am doing something wrong here, but the option
'continuous_Bayesian' seems currently not usable to me.
Regards
Juliane