spark-instrumented-optimizer/python/pyspark/sql
ksonj 3df9c5dd0c Better error message on access to non-existing attribute
I believe column access via `__getattr__` is bad and shouldn't be implicitly encouraged by the error message when accessing a non-existing attribute on DataFrame. This patch changes the error message from 'no such column'  to the more generic 'no such attribute', which is also what Pandas DFs will throw.

Author: ksonj <kson@siberie.de>

Closes #5771 from ksonj/master and squashes the following commits:

bcc2220 [ksonj] Better error message on access to non-existing attribute
2015-04-29 09:48:47 -07:00
..
__init__.py [SPARK-4897] [PySpark] Python 3 support 2015-04-16 16:20:57 -07:00
_types.py [SPARK-6949] [SQL] [PySpark] Support Date/Timestamp in Column expression 2015-04-21 00:08:18 -07:00
context.py [SPARK-6949] [SQL] [PySpark] Support Date/Timestamp in Column expression 2015-04-21 00:08:18 -07:00
dataframe.py Better error message on access to non-existing attribute 2015-04-29 09:48:47 -07:00
functions.py [SPARK-7188] added python support for math DataFrame functions 2015-04-29 00:09:24 -07:00
mathfunctions.py [SPARK-7188] added python support for math DataFrame functions 2015-04-29 00:09:24 -07:00
tests.py [SPARK-7188] added python support for math DataFrame functions 2015-04-29 00:09:24 -07:00