[madlib] branch master updated: MLP: Fix column names in prediction output table

[njayaram]

This is an automated email from the ASF dual-hosted git repository.

njayaram pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/madlib.git


The following commit(s) were added to refs/heads/master by this push:
     new 01bfbcb  MLP: Fix column names in prediction output table
01bfbcb is described below

commit 01bfbcbfb4547bc0e60f2e9a315f8bcbe9ec66e1
Author: Himanshu Pandey <hpan...@pivotal.io>
AuthorDate: Thu May 2 13:44:55 2019 -0700

    MLP: Fix column names in prediction output table
    
    JIRA: MADLIB-1323
    Change column names for pred_type=prob to have 'prob_' prefix, instead
    of 'estimated_prob_'. This commit also adds test cases for this change
    in dev-check.
    Closes #384
---
 src/ports/postgres/modules/convex/mlp_igd.py_in   | 12 ++++--------
 src/ports/postgres/modules/convex/test/mlp.sql_in |  7 +++++++
 2 files changed, 11 insertions(+), 8 deletions(-)

diff --git a/src/ports/postgres/modules/convex/mlp_igd.py_in 
b/src/ports/postgres/modules/convex/mlp_igd.py_in
index 3d93a9a..b290c5f 100644
--- a/src/ports/postgres/modules/convex/mlp_igd.py_in
+++ b/src/ports/postgres/modules/convex/mlp_igd.py_in
@@ -59,6 +59,8 @@ from utilities.validate_args import output_tbl_valid
 from utilities.validate_args import table_exists
 from utilities.validate_args import quote_ident
 from utilities.minibatch_validation import validate_dependent_var_for_minibatch
+from utilities.utilities import create_cols_from_array_sql_string
+
 
 
 @MinWarning("error")
@@ -1048,14 +1050,8 @@ def mlp_predict(schema_madlib, model_table, data_table, 
id_col_name,
         else:
             intermediate_col = unique_string()
             if classes:
-                # we cannot call sql quote_ident on the class value because
-                # aliasing does not support quote_ident. Hence calling our
-                # python implementation of quote_ident
-                score_format = ',\n'.join([
-                    'CAST({interim}[{j}] as DOUBLE PRECISION) as "{c_str}"'.
-                    format(j=i + 1, c_str=quote_ident("estimated_prob_{0}".
-                        format(str(c))).strip(' "'), interim=intermediate_col)
-                    for i, c in enumerate(classes)])
+                score_format = create_cols_from_array_sql_string(
+                    classes, intermediate_col, 'prob', 'double precision', 
False, 'MLP') 
             else:
                 # Case when the training step did not have to one-hot encode
                 # the dependent var.
diff --git a/src/ports/postgres/modules/convex/test/mlp.sql_in 
b/src/ports/postgres/modules/convex/test/mlp.sql_in
index 2d66815..9053df7 100644
--- a/src/ports/postgres/modules/convex/test/mlp.sql_in
+++ b/src/ports/postgres/modules/convex/test/mlp.sql_in
@@ -454,6 +454,13 @@ SELECT mlp_predict(
     'mlp_prediction_batch_output',
     'prob');
 SELECT * FROM mlp_prediction_batch_output;
+
+-- assert to test if the newly created column name exists and matches the count
+SELECT assert(count(*) =3, 'Column Names does not exists')
+          from pg_attribute
+          where attrelid = 'mlp_prediction_batch_output'::regclass
+       and attname in( 'prob_1', 'prob_2','prob_3');
+
 DROP TABLE IF EXISTS mlp_prediction_batch_output;
 DROP TABLE IF EXISTS mlp_class_batch, mlp_class_batch_summary, 
mlp_class_batch_standardization;