PMMLlib Python examples

Update a model in an existing PMML file :

The updating is done in two steps:

• 1 : delete the XML node of the model with method UnlinkNode();
• 2 : re-create the model.

  PMMLlib p ( fileName, log );
  
  # Set the model
  p.SetCurrentModel( modelName, modelType );
  
  # Delete the XML node of the model
  p.UnlinkNode( );
  # Recreate the model with new parameters
  p.AddRegressionModel(« monModele », PMMLlib::kREGRESSION,  « regression » );
  p.AddDataField( ….);
  
  # Save the PMML file
  p.Write( );
  

Backup and update a model in an existing PMML file :

It is done in two steps:

• 1 : backup the model in an XML node with name modelName_ with method BackupNode();
• 2 : re-create the model.

  PMMLlib p ( fileName, log );
  
  # Set the model
  p.SetCurrentModel( « monModele », modelType );
  
  # Save the model in a new XML node
  p.BackupNode( );
  # Modify
  p.AddRegressionModel(« monModele », PMMLlib::kREGRESSION,  « regression » );
  p.AddDataField( ….);
  
  # Save the PMML file
  p.Write( );
  

Add a model in an existing PMML file :

PMMLlib p ( fileName, log );

# Create the model
p.AddRegressionModel(« monModele », PMMLlib::kREGRESSION,  « regression » );
p.AddDataField( ….);

# Save the PMML file
p.Write( );

Read a model and execute it :

P = PMMLlib( fileName, log );
p.SetCurrentModel( modelName, modelType );

pyStrCode = p.ExportPythonStr( « myPyFunc », « function header » ); 
exec pyStrCode;

# Eval myPyFunc which is now known as a python function
inputs = [1.,2.,3.,4.]
res = myPyFunc(inputs)