Re: annual average from monthly mean for precise 365-day years.

[Wei Cheng <wcheng@xxxxxxxxxxxxxxxxxxxx>]

> The best/right way to perform annual averaging is to define the desired
> result axis so its cell boundaries exactly align with the cells of the source
> axis. Your Method 2 looks like this approach. I have modified your Method 2
> just slightly and plotted it using the V5.3 PLOT/STEP style to illustrate
>
>      define axis/edges/t=1:25:1/unit=months tax
>      let newtaux = T[GT=tax]
>      plot/step newtaux
>      define axis/edges/t=1:25:12/UNITS=months tann
>      let tauxx=newtaux[gt=tann]
>      plot/step/over tauxx

Hi, Steve,

Thanks very much! Still, I have one last problem:

In your above example, if number "25" is replaced by "24",
and, if line

"let tauxx=newtaux[gt=tann]"

is replaced by

"let tauxx=newtaux[gt=tann@ave]"

By "list tauxx", we can see that the last point in tauxx is a bad value.
This must be due to using @ave, plus the fact that now the total data
point is an even number (24 instead of 25).

How can I get the last point in tauxx back (it should still exist)?
In climate studies, number of months is even for multiple whole
year's data (how unlucky it is that each year has 12 months). But
lucky, we have you!

Thanks again.

Wei