Details
-
Bug
-
Status: Closed
-
Major
-
Resolution: Fixed
-
None
-
None
Description
Decision tree install-check fails intermittently when run on a big, distributed GPDB server with below error:
DROP TABLE IF EXISTS train_output, train_output_summary; psql:/tmp/madlib.xStCxm/recursive_partitioning/test/decision_tree.sql_in.tmp:273: NOTICE: table "train_output" does not exist, skipping psql:/tmp/madlib.xStCxm/recursive_partitioning/test/decision_tree.sql_in.tmp:273: NOTICE: table "train_output_summary" does not exist, skipping DROP TABLE SELECT tree_train('dt_golf'::text, -- source table 'train_output'::text, -- output model table 'id'::text, -- id column 'temperature::double precision'::text, -- response 'humidity, windy, "Cont_features"'::text, -- features NULL::text, -- exclude columns 'gini'::text, -- split criterion NULL::text, -- no grouping NULL::text, -- no weights 10::integer, -- max depth 6::integer, -- min split 2::integer, -- min bucket 8::integer, -- number of bins per continuous variable 'cp=0.01' -- cost-complexity pruning parameter ); psql:/tmp/madlib.xStCxm/recursive_partitioning/test/decision_tree.sql_in.tmp:288: ERROR: plpy.SPIError: Function "_dst_compute_entropy_transition(integer[],integer,integer)": Invalid type conversion. Null where not expected. (seg62 slice1 sdw8.gphd.local:1031 pid=457374) (plpython.c:4656) CONTEXT: Traceback (most recent call last): PL/Python function "tree_train", line 28, in <module> null_handling_params, verbose_mode) PL/Python function "tree_train", line 484, in tree_train PL/Python function "tree_train", line 264, in _get_tree_states PL/Python function "tree_train", line 720, in _get_bins PL/Python function "tree_train"
Attachments
Issue Links
- links to