I have a problem in running weka training and predictor. My traning set has 456 compounds(rows) and 156 descriptors (columns) and one Column with classes. Test set has 6650 compounds(rows) and 156 descriptors(Columns). when i run the node, it shows following error
Failed to apply settings: java.lang.OutOfMemoryError: Java heap space
Could you please help me in this...
Make sure the structures are in smiles format and not sdf, chemaxon or something similar
I dont have smiles in the node. I have filtered those structures before i go to model development (AttributeSelectedClassifier).
In the config window go to memory control and select option "Write table to disc" to see if that helps.
Another option is to edit the knime.ini file, and then reload KNIME with the settings (it may be the Java memory was set to low):