This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Greetings Akshay, I think the issue is that you are extracting the date from the filename (good), but are rewriting the information using the NCML ouside of the aggregation (i.e. <values start="0" increment="1" />). Can you try using the same NcML you have, but remove this part of the NcML: <variable name="TIME" type="double" shape="TIME"> <attribute name="units" value="months since 1960-01-01 00:00:00" /> <values start="0" increment="1" /> <attribute name="axis" value="T" /> </variable> Thanks! Sean address@hidden> wrote: > New Ticket: Regarding NCML Aggregating ( montly axis at fixed date ) > > Hi! > > This query is regarding time variable declaration in NCML, here my aim > is to define time axis at 1th of every month that is same as that of > date stamp in filename. > > > My Input : > > I have list of files like below > > # ls -1 | head -10 > air_19600101.nc > air_19600201.nc > air_19600301.nc > air_19600401.nc > air_19600501.nc > air_19600601.nc > air_19600701.nc > air_19600801.nc > air_19600901.nc > air_19601001.nc > > > This is what I tried : > > Following is my ncml > > <netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2"> > <variable name="TIME" type="double" shape="TIME"> > <attribute name="units" value="months since 1960-01-01 00:00:00" /> > <values start="0" increment="1" /> > <attribute name="axis" value="T" /> > </variable> > <aggregation dimName="TIME" type="joinExisting"> > <scan dateFormatMark="air_#yyyymmdd" location="/path/to/AIR" > suffix=".nc" subdirs="false" /> > </aggregation> > </netcdf> > > > This is what I received after restarting thredds data server: > > I use ferret for data analysis, this is how time indexes are created > > yes? set data "http://127.0.0.1/thredds/dodsC/test/test.nc" > yes? sh g air > GRID GFP1 > name axis # pts start end > LON LONGITUDE 120mr 20.5E 139.5E > LAT LATITUDE 115 r 74.5S 39.5N > normal Z > TIME TIME 674 r 01-JAN-1960 12:00 31-JAN-2016 > 12:24 > normal E > normal F > yes? list/l=1:10 t[gt=air] > VARIABLE : T > axis TIME > FILENAME : test.nc > FILEPATH : http://127.0.0.1/thredds/dodsC/test/ > SUBSET : 10 points (TIME) > 01-JAN-1960 / 1: 0.000 > 31-JAN-1960 / 2: 1.000 > 02-MAR-1960 / 3: 2.000 > 01-APR-1960 / 4: 3.000 > 02-MAY-1960 / 5: 4.000 > 01-JUN-1960 / 6: 5.000 > 02-JUL-1960 / 7: 6.000 > 01-AUG-1960 / 8: 7.000 > 31-AUG-1960 / 9: 8.000 > 01-OCT-1960 / 10: 9.000 > > > This is expected output : > > Here I expect time index to be created like below, at first of every > month, or else 15th of every month > > 01-JAN-1960 / 1: 0.000 > 01-FEB-1960 / 2: 1.000 > 01-MAR-1960 / 3: 2.000 > 01-APR-1960 / 4: 3.000 > 01-MAY-1960 / 5: 4.000 > 01-JUN-1960 / 6: 5.000 > 01-JUL-1960 / 7: 6.000 > 01-AUG-1960 / 8: 7.000 > 01-SEP-1960 / 9: 8.000 > 01-OCT-1960 / 10: 9.000 > > > Hope someone can guide me > > Thanks in advance. > > Have a nice day ! > > Regards, > Akshay Hegde > > > Ticket Details > =================== > Ticket ID: UXM-196070 > Department: Support netCDF Java > Priority: Normal > Status: Open > Link: > https://andy.unidata.ucar.edu/esupport/staff/index.php?_m=tickets&_a=viewticket&ticketid=26875 > > Ticket Details =================== Ticket ID: UXM-196070 Department: Support netCDF Java Priority: Normal Status: Open