Changed the data type of the 'number of rows' column in the summary table of linear regression training from integer to BIGINT, in order to prevent an integer out of range exception in cae the total number of rows in the input table exceeds the range of a signed integer
diff --git a/src/ports/postgres/modules/regress/linear.py_in b/src/ports/postgres/modules/regress/linear.py_in
index 2be78e9..364b380 100644
--- a/src/ports/postgres/modules/regress/linear.py_in
+++ b/src/ports/postgres/modules/regress/linear.py_in
@@ -180,7 +180,7 @@
, '{out_table}'::varchar as out_table
, $__madlib__${dependent_varname}$__madlib__$::varchar as dependent_varname
, $__madlib__${independent_varname}$__madlib__$::varchar as independent_varname
- , {num_rows_processed}::integer as num_rows_processed
+ , {num_rows_processed}::BIGINT as num_rows_processed
, {num_rows_skipped}::integer as num_missing_rows_skipped
, {grouping_col}::text as grouping_col
""".format(out_table=out_table, source_table=source_table,
