Re: 20040222: Suitability of NetCDF to Telemetry

>To: address@hidden
>From: "John R. Cox" <address@hidden>
>Subject: Suitability of NetCDF to Telemetry
>Organization: ?
>Keywords: 200402222329.i1MNTPrV007651 netCDF telemetry data

Hi John,

> I'm trying to figure out whether NetCDF is suitable for use with Pulse 
> Code Modulated (PCM) data that is supercommutated or subcommutated 
> without reformatting the data into a simpler structure.
> 
> It appears as if data can be accessed with a "stride" which might make 
> it suitable for subcom, but it doesn't appear that variables can be 
> defined with a built-in stride.
> 
> Is that so?

You could define a variable attribute, say "stride", that had the
value of the stride you want, but this would just be a convention.  If
you wrote a layer on top of the netCDF nc_get_vars_TYPE() call that
honored the value of this stride attribute when it existed, then I
think you would have what you want.  You would just have to make sure
code that accessed the variable used your function rather than the
library function.

Or am I misunderstanding your question?

--Russ