Re: UNIT8 and floating point data

• To: netcdfgroup@xxxxxxxxxxxxxxxx
• Subject: Re: UNIT8 and floating point data
• From: PEPKE@xxxxxxxxxxxx
• Date: Fri, 18 Feb 1994 14:27:15 -0500 (EST)

Russ writes:

> One way to do this is to pack floating-point numbers into ncbyte or ncshort
> values and use the conventional netCDF attributes `scale_factor' and
> `add_offset' to store the packing parameters, as described in the User's
> Guide:

Thank you.  That seems to be the correct way to do it in a standard manner,
when the mapping is a linear equation.  My brain must have been off when I 
was reading that part of the manual.  Of course, the missing value can be 
compared before the scaling.

It doesn't directly address the problem of NEXRAD-type data where it comes in
with the table information built in.  However, I think that case will be 
unusual enough so that having an ad-hoc indirection to a table won't cause 
that much of a problem.  Internally I'm going to use a table in any event,
because it's fast.

Eric Pepke                                     INTERNET: pepke@xxxxxxxxxxxx
Supercomputer Computations Research Institute  MFENET:   pepke@fsu
Florida State University                       SPAN:     scri::pepke
Tallahassee, FL 32306-4052                     BITNET:   pepke@fsu

Disclaimer: My employers seldom even LISTEN to my opinions.
Meta-disclaimer: Any society that needs disclaimers has too many lawyers.