Closed mhidas closed 6 years ago
The specific issues the checker complains about are as below.
§2.6.2 Recommended Global Attributes :2: 1/ 2 : global attribute history
should exist and be a non-
empty string
This is easy to fix.
§2.4 Dimension Order :2: 4/ 7 : PRES's dimensions are not
in the recommended order
T, Z, Y, X. They are PRES,
INSTANCE, TEMP's
dimensions are not in the
recommended order T, Z, Y,
X. They are PRES,
INSTANCE, PSAL's
dimensions are not in the
recommended order T, Z, Y,
X. They are PRES, INSTANCE
§5.0 multidimensional coordinate PRES s:2: 0/ 1 : PRES shares the same name
as one of its dimensions
The easiest way to fix both of these issues is to remove the INSTANCE
dimension (leaving the variable). Since it has a length of 1, it's not needed.
Alternatively, if you want to keep this dimension, allowing for the possibility of it being longer than 1, it needs to be the first dimension of all those variables, before any space or time dimensions (see CF for more info). In that case, however, the PRES
dimension needs to be named differently, e.g. LEVEL
or Z
.
§2.4 Dimension Order :2: 9/11 : TEMP's dimensions are not
in the recommended order
T, Z, Y, X. They are
DEPTH, INSTANCE,
TEMP_quality_control's
dimensions are not in the
recommended order T, Z, Y,
X. They are DEPTH,
INSTANCE
§2.6.2 Recommended Attributes :2: 4/ 5 : institution should be
defined
The first of these is the same as for AATAMS (remove INSTANCE dimension, or switch order). The second is simply a missing global attribute.
@lbesnard @bpasquer @xhoenner I'm happy to fix up the test files, but the above fixes will also need to be applied to the actual data files coming through, otherwise they will all fail the CF checks.
Fixed in #17
We now have three unittests failing due to test files that were expected to pass the checks, but now fail with checker version 3.