On 3/18/11 12:21 PM, John Caron wrote:
CF has the calendar attribute (see section 4.4.1)
and thanks for posting that -- so that's taken care of.
well, plenty of algorithms just want, say, the monthly average. but if
you want to know the exact time interval, you need a library that deals
with calendars. the nice thing about the current udunits way of doing
things is that the data provider calculates the interval for you.
unfortunately, for complicated cases, she may not do it correctly (so we
should create a standard library that does it), and/or the calendar date
that you derive from the time unit may not be what was intended (so we
should create a standard library that does it)
I'm all for libraries, but it seems like code for "calculate the monthly
average" of a given data set has nothing to do with CF, and is getting
beyond what udunits should do, and probably beyond what a reference
library associated with CF should be.
That would be in a lib for doing met/ocean time series analysis would
have in it. I"m sure there are a number of these.
Here's one I know of, but haven't used:
http://pytseries.sourceforge.net/
Right there in the home page is says: "Convert a daily time series to
monthly by taking the average value during each month"
So what CF needs is to specify the data in a way that a lib like that
can use it. It seems we are approaching that.
-Chris
--
Christopher Barker, Ph.D.
Oceanographer
Emergency Response Division
NOAA/NOS/OR&R (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception
[email protected]
_______________________________________________
CF-metadata mailing list
[email protected]
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
