Github user njayaram2 commented on a diff in the pull request:
https://github.com/apache/incubator-madlib/pull/44#discussion_r65597573
--- Diff: src/ports/postgres/modules/utilities/sessionize.py_in ---
@@ -0,0 +1,101 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+import plpy
+import string
+
+from control import MinWarning
+from utilities import unique_string, _assert
+from validate_args import get_cols
+from validate_args import input_tbl_valid, output_tbl_valid, is_var_valid
+
+m4_changequote(`<!', `!>')
+
+def sessionize(schema_madlib, source_table, output_table, partition_expr,
+ time_stamp, time_out, **kwargs):
+ """
+ Perform sessionization over a sequence of rows.
+
+ Args:
+ @param schema_madlib: str, Name of the MADlib schema
+ @param source_table: str, Name of the input table/view
+ @param output_table: str, Name of the table to store result
+ @param partition_expr: str, Expression to partition (group) the
input data
+ @param time_stamp: float, Column name with time used for
sessionization calculation
+ @param time_out: float, Delta time between subsequent events to
define a sessions
+
+ """
+ with MinWarning("error"):
+ _validate(source_table, output_table, partition_expr,
time_stamp, time_out)
+
+ all_input_cols_str = ', '.join([i.strip() for i in
get_cols(source_table, schema_madlib)])
+ session_id = 'session_id' if not is_var_valid(source_table,
'session_id') else unique_string('session_id')
+
+ plpy.execute("""
+ CREATE TABLE {output_table} AS
+ SELECT
+ {all_input_cols_str},
+ CASE WHEN {time_stamp} NOTNULL
+ THEN (SUM(new_event_boundary)
OVER (PARTITION BY {partition_expr} ORDER BY {time_stamp})) END AS {session_id}
+ FROM (
+ SELECT *,
+ CASE WHEN {time_stamp}
NOTNULL and ({time_stamp}-LAG({time_stamp},1) OVER (w) > '{time_out}' OR
ROW_NUMBER() OVER (w) = '1')
--- End diff --
_Code-wise, I think the best way to handle this is to see if the python
interval parameter is some kind of number and if it is pass it directly
to Postgres and multiply it by a 1 second interval, ie: "{time_out} *
interval '1 second'". Postgres will do the correct thing regardless of
what numeric type it is (int vs numeric vs float). Otherwise just assume
it's something that can be cast to an interval and do
"{time_out}::interval"._
If I understood you correctly, {time_out} and {min_time} would still be
strings. They can
either have a value such as '30', in which case we would internally convert
it to 30 seconds,
or it can have a string with non-numeric characters which will be cast to
type interval
(the casting will fail if the string is not in line with interval's format).
But, Postgres' interval type seems to consider something like '30' to be a
legitimate interval
value (I reckon it is considered as 30 hours). Will have to check if
converting '30' to seconds
instead of letting {time_out}::interval handle it is acceptable from the
API's perspective.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
