Re: [ferret_users] calcuting RMSE

[Nicolas Freychet <nicolas.freychet@xxxxxxxxx>]

Hi Juliet,

It just seems that your 2 datasets do not have the same calendar or time record. If they only have 1 timestep, you can use the command "L=1" to precise that each variable should be associated to its first time:

let p=pre[d=2,L=1]-pr[d=1,L=1]

(if they have more than one timestep you may need to precise the time you consider differently, for example averaging the 6 first times L=1:6@ave.....)

2016-10-10 15:35 GMT+01:00 afwande juliet <afwandej965@xxxxxxxxx>:

Dear ALL, I am trying to calculate RMSE for two datasets over a box for MAM seasons for 1997 only. I have re-gridded the data so that have same grid information. But I get the error below. Could some help me. Also attached are two datasets am using

yes?  set region/x=32E:40E/y=8N:18S
yes? let p=pre[d=2]-pr[d=1]
yes? let q = p^2
yes? let r = q[l=1:3@sum]
yes? let s = r/3
yes? let m = s^0.5
yes? list m
           *** NOTE: Ambiguous coordinates on T axis: PRE[D=2]-PR[D=1]
 **ERROR: illegal limits: PR is not in the range T=35488:35579
          Axis extremes are T=19970301:19970532
yes?  set region/x=32E:40E/y=8N:18S
yes? let p=pre[d=2]-pr[d=1]
yes? let q = p^2
yes? let r = q[l=1:3@sum]
yes? let s = r/3
yes? let m = s^0.5
yes? list m
           *** NOTE: Ambiguous coordinates on T axis: PRE[D=2]-PR[D=1]
 **ERROR: illegal limits: PR does not exist at T=19970516
          Axis extremes are T=19970516

--

----------------------------------------------------

Nicolas Freychet

PDRA, School of Geosciences
University of Edinburgh
----------------------------------------------------