Error with data set matching for parameter learning

The engine.
Post Reply
EAly
Posts: 21
Joined: Tue Apr 14, 2020 1:16 pm

Error with data set matching for parameter learning

Post by EAly »

Hello,
I'm trying to perform parameters learning for a network I created using pysmile, however, I get the following error when trying to

Code: Select all

 match_network 

Code: Select all

SMILEException: SMILE Error Occured in: DataSet.MatchNetwork Dataset column 'ind_1_5_1' is continous, but matching node is discrete
ErrNo=-51
So as far as I understand, the possible solutions for this error is either to change the type of the node or do some data manipulation -although I'm not what change do I need to do to the data-. So if I'm to change the type of nodes, can I do that after creating them, something like "set_node_type"? and can I create an equation node and do the learning without defining a node equation?
I would appreaciate if you can also suggest a better solution?

Thanks in advance for your help

PS. Attached is the datafile, but couldn't attach the network .xdsl file because of its big size
Attachments
All_variables.csv
Data file
(85.65 KiB) Downloaded 209 times
shooltz[BayesFusion]
Site Admin
Posts: 1417
Joined: Mon Nov 26, 2007 5:51 pm

Re: Error with data set matching for parameter learning

Post by shooltz[BayesFusion] »

You need to discretize your data before parameter learning if your nodes are discrete. ind_1_5_1 contains values like 46.29 - SMILE can't map this onto a discrete node.
EAly
Posts: 21
Joined: Tue Apr 14, 2020 1:16 pm

Re: Error with data set matching for parameter learning

Post by EAly »

Thanks for your reply Shooltz,
but is there an option to do it the other way around, i.e. make the node continues without defining an equation?
shooltz[BayesFusion]
Site Admin
Posts: 1417
Joined: Mon Nov 26, 2007 5:51 pm

Re: Error with data set matching for parameter learning

Post by shooltz[BayesFusion] »

No, the continuous nodes in GeNIe/SMILE are always defined by the equation.
Post Reply