ow to set the time variables from 0,1,2,3,4,5 into yyyy-mm-dd, hh-mm-ss
Added by Sam United almost 7 years ago
Hi all,
Now I have a nc file with monthly data and the 'time' variable is: "time= 0,1,2,3,4,5 "but I want to change it into a nc file and "time = 1997-01-01 00:00:00, 1997-02-01 00:00:00, 1997-03-01 00:00:00, 1997-01-04 00:00:00, 1997-05-01 00:00:00". How I can do this. perhaps using setdate? but now sure.
Cheers
Sam
Replies (6)
RE: ow to set the time variables from 0,1,2,3,4,5 into yyyy-mm-dd, hh-mm-ss - Added by Ralf Mueller almost 7 years ago
Hi Sam!
check the operator settaxis
!
cdo -settaxis,1997-01-01,00:00:00,1months <ifile> <ofile>should do it
cheers
ralf
RE: ow to set the time variables from 0,1,2,3,4,5 into yyyy-mm-dd, hh-mm-ss - Added by Ralf Mueller almost 7 years ago
and please update CDO! 1.6.3 is already petrified .... ;-)
RE: ow to set the time variables from 0,1,2,3,4,5 into yyyy-mm-dd, hh-mm-ss - Added by Sam United almost 7 years ago
Hi Ralf,
Thank you for your help, but unfortunately, it failed to do that although the unit of 'time' is changed into 'months from 1997.01.01.00.00.00', the data of 'time' is still like 0,1,2,3,4,... And the picture of the output file is as attachment (file is too large to upload)
Cheers
Sam
RE: ow to set the time variables from 0,1,2,3,4,5 into yyyy-mm-dd, hh-mm-ss - Added by Ralf Mueller almost 7 years ago
and these values are correct. 0,1,2 ... months since 1970-01-01. cf convention does not allow to save time as a string afaik. just like any other variable, time valuea have to be correctly interpreted according to thwir given unit. please, check the file with the infov
operator.
RE: ow to set the time variables from 0,1,2,3,4,5 into yyyy-mm-dd, hh-mm-ss - Added by Sam United almost 7 years ago
Hi Ralf,
Thanks, yes the nc file is created by python, are there some ways now to help me to do that?
RE: ow to set the time variables from 0,1,2,3,4,5 into yyyy-mm-dd, hh-mm-ss - Added by Ralf Mueller almost 7 years ago
could you upload the file?
as I said: I might be technically possible to save timesteps as strings ("1997.01.01.00.00.00"), but I really dont recommend that since
- you cannot do any computation with them
- the file does not follow the CF-convention (http://cfconventions.org/Data/cf-conventions/cf-conventions-1.7/cf-conventions.html#time-coordinate), hence you will have a hard time with tools relying on it (like CDO).
hth
ralf