Examples | Logistic Regression Scoring | Vantage Analytics Library - Examples - Vantage Analytics Library

Logistic Regression that Subsequent Examples Score

call td_analyze (
  'logistic',
  'database = val_source;
   tablename = customer;
   columns = income, age, years_with_bank;
   outputdatabase = val_results;
   outputtablename = log_example;
   dependent = nbr_children;
   response = 0;'
);

Scoring without Model Evaluation

call td_analyze (
  'logisticscore',
  'database = val_source;
   tablename = customer;
   modeldatabase = val_results;
   modeltablename = log_example;
   outputdatabase = val_results;
   outputtablename = log_example1;
   probability = Probability;
   scoringmethod = score;'
);

Model Evaluation without Scoring

Creates Success Table, Multi-Threshold Success Table and Lift Table by default.

call td_analyze (
  'logisticscore',
  'database = val_source;
   tablename = customer;
   modeldatabase = val_results;
   modeltablename = log_example;
   outputdatabase = val_results;
   outputtablename = log_example2;
   estimate = Estimate;
   scoringmethod = evaluate;'
);

Scoring and Model Evaluation

call td_analyze (
  'logisticscore',
  'database = val_source;
   tablename = customer;
   modeldatabase = val_results;
   modeltablename = log_example;
   outputdatabase = val_results;
   outputtablename = log_example3;
   estimate = Estimate;
   scoringmethod = scoreandevaluate;'
);

Specify Various Options

call td_analyze (
  'logisticscore',
  'database = val_source;
   tablename = customer;
   modeldatabase = val_results;
   modeltablename = log_example;
   outputdatabase = val_results;
   outputtablename = log_example4;
   estimate = Estimate;
   probability = Probability;
   retain = state_code;
   samplescoresize = 10;
   scoringmethod = scoreandevaluate;
   lifttable = true;'
);