Github user iyerr3 commented on a diff in the pull request:
https://github.com/apache/madlib/pull/295#discussion_r203811750
--- Diff:
src/ports/postgres/modules/recursive_partitioning/decision_tree.py_in ---
@@ -2327,6 +2328,110 @@ def _tree_error(schema_madlib, source_table,
dependent_varname,
plpy.execute(sql)
# ------------------------------------------------------------
+def _validate_var_importance_input(model_table, summary_table,
output_table):
+ _assert(table_exists(model_table),
+ "Recursive Partitioning: Model table does not exist.")
+ _assert(table_exists(summary_table),
+ "Recursive Partitioning: Model summary table does not exist.")
+ _assert(not table_exists(output_table),
+ "Recursive Partitioning: Output table already exists.")
+
+def _is_model_for_RF(summary_table):
+ # Only an RF model (and not DT) would have num_trees column in summary
+ return columns_exist_in_table(summary_table, ['num_trees'])
--- End diff --
We've a `method_name` in summary table that makes this clearer.
---
