Russ Rew: Re: 970127: records

Oops, forgot to CC: support-netcdf on this reply ...

------- Forwarded Message

Date:    Mon, 27 Jan 1997 10:24:26 -0700
From:    Russ Rew <address@hidden>
To:      address@hidden (Bjorn Stevens)
Subject: Re: 970127: records 

>To: address@hidden
>From: address@hidden (Bjorn Stevens)
>Subject: records
>Organization: .
>Keywords: 199701271628.JAA22433

Bjorn,

> Thanks for the help.  I'll give it a go.  I think my problem before
> was that for the nth time I was setting count(4) and start(4) = n,
> upon rereading the manual and your example i guess i misinterpreted
> the meaning of count.
> 
> Thanks again,
> 
> Bjorn
> 
> PS I'm wrinting in fortran so the outer dimension (4) is my slowest
> varying one, hence the difference with respect to your example.

Yes, sorry, my example was in error.  I forgot that the FORTRAN
interface reverses the values of the start/count dimensions to
correspond to the Fortran-order of dimensions.  The example should be:

    then you can put a whole time's worth of data with a single call to
    ncvpt:

        integer start(4), count(4)
        data start /1, 1, 1, 1/
        data count /LON, LAT, LEVEL, 1/
          ...
        ! to write the nth time
        start(4) = n
        call ncvpt(ncid, varid, start, count, var, rcode)

I hope I got it right this time!  I can certainly understand that this
is confusing ...

- --Russ


------- End of Forwarded Message