Re: [ferret_users] how to integrate within a mixed layer?

[Peter Szabo <szabpet83@xxxxxxxxx>]

Hi Yangxing,

you want to use the DIN function, when "end" not equals missing values, do not you?
then (there was also a typing mistake in your script, but) try this:

let missing = `mld[d=2],r=bad`;\

let heat_mld = if end ne missing then heat[d=1,z=`beg`:`end`@din];\

 the "else" is needless at the end of "if".

Peter

On Thu, May 29, 2008 at 10:25 PM, yangxing zheng <yangxingz@xxxxxxxxx> wrote:

Hi Peter,

   Thanks for great help. I still have questions.
Since there is missing values in land for ocean MLD
datasets, I try using mask to define integration

repeat/l=1:10 (repeat/j=1:100 (repeat/i=1:100 (let
end=mld[d=2];\
let heat_mld = if end ne (0-(1e+34)) then
heat[d=1,z=`beg`:`end`@din else (0-(1e+34));\
save/file="heat_mld.nc"/append/ilimits=1:100/jlimits=1:100/llimits=1:10/i=`i`/j=`j`/l=`l`/clobber
heat_mld)))

....

The major problem is the definition of
heat[d=1,z=0:bad@din] fails as end = missing values

How to avoid something like z=0:bad@din?