Problem Note 67015: An incorrect format might be added to a column that is defined as a BOOLEANÂ or a NUMERIC data type in the Google BigQuery database
Reading in a Google BigQuery table that contains columns that are defined as a BOOLEAN or NUMERIC data type might result in the wrong format being assigned to those columns. This issue occurs under the following conditions:
- When using explicit pass-through to read in the table, a Boolean column is assigned a 1. format rather than the documented 4. format. The documentation is incorrect, and the 1. format is the correct format.
- When using explicit pass-through to read in the table, a numeric column is assigned a dollar. format instead of the documented w.9 format. The expected behavior is that no format be applied to the variable
- When using a LIBNAME statement to read in the table, the numeric column is not assigned any format instead of the w.9 format. The documentation is incorrect, and the 1. format is the correct format.
The workaround is to use a LIBNAME statement rather than using explicit pass-through to read in the BigQuery table. You can also add the format explicitly rather than letting it default.
Click the Hot Fix tab in this note to access the hot fix for this issue.
Applying this hot fix changes the default SAS format assigned for the following BigQuery data types:
Data Type |
Prior Documented Format |
New Expected Format |
BOOLEAN |
4. |
1. |
FLOAT64 |
w.d |
none |
NUMERIC |
w.9 |
none |
Operating System and Release Information
SAS System | SAS/ACCESS Interface to Google BigQuery (on SAS Viya) | Linux for x64 | V.03.04 | 2020.1.5 | Viya | Viya |
SAS System | SAS/ACCESS Interface to Google BigQuery | Linux for x64 | 9.4 | 9.43 | 9.4 TS1M6 | 9.4 TS1M8 |
*
For software releases that are not yet generally available, the Fixed
Release is the software release in which the problem is planned to be
fixed.
Type: | Problem Note |
Priority: | high |
Date Modified: | 2020-12-07 11:15:31 |
Date Created: | 2020-12-01 14:43:13 |