Seth,
On Wed, May 22, 2013 at 5:04 PM, Seth McGinnis mcgin...@ucar.edu wrote:
Hi Ellyn,
According to CF Trac Ticket #31 (slated for inclusion in the update to CF
1.7),
the way to cache minimum maximum values in metadata is to use an attribute
named actual_range and store them as a pair.
Seth-
Thanks very much for the information and pointer to the ticket
documenting the issue!
Do you know if NaN is an allowed value for actual_range? In our
historical data, a variable is retained if it was collected, even though
the sensor failed completely (and all the data was replaced
Ellyn,I believe that the valid_range and actual_range attributes are intended to only list non-fill values. The actual range should fall inside the valid range, and both should exclude invalid values.Grace and peace,Jim
Jim BiardResearch ScholarCooperative Institute for Climate and
Computing the min max on the fly is cheap, and approximating it is even
cheaper, so why introduce the uncertainty?
... but computing min max on the fly can also be very expensive.
We have aggregated model output datasets where each variable is more
than 1TB!
Sure, I can see that that's
Seth,
On Thu, May 23, 2013 at 9:51 AM, Seth McGinnis mcgin...@ucar.edu wrote:
Computing the min max on the fly is cheap, and approximating it is even
cheaper, so why introduce the uncertainty?
... but computing min max on the fly can also be very expensive.
We have aggregated model output
Thank you Jim!
When we transition our older files to CF, I'll have to leave out
the variables for which there is no good data.
Best, Ellyn
On 05/23/2013 08:49 AM, Jim Biard wrote:
Ellyn,
I believe that the
... but computing min max on the fly can also be very expensive.
We have aggregated model output datasets where each variable is more
than 1TB!
Sure, I can see that that's useful metadata about the dataset, and that
there's value in caching it somewhere. I just don't think it belongs with
+1 Martin. I am bugged (not the technical term) by the conclusions here, which
seem to be: Because people design systems badly, I must constrain my own system
to accommodate their failures.
The use cases for storing the summary information with the file are: (A) It's
faster to access, which
Well said, John!There are many good reasons for storing discovery and provenance metadata within our netCDF files. You can make an argument for storing such metadata only elsewhere, but I think this is only justified if you have a publicly accessible database system in place to serve the
On Thu, May 23, 2013 at 9:00 AM, John Graybeal
grayb...@marinemetadata.org wrote:
+1 Martin. I am bugged (not the technical term) by the conclusions here,
which seem to be: Because people design systems badly, I must constrain my
own system to accommodate their failures.
The use cases for
Folks-
I have a two-parter for your consideration today.
Is there a recommended best practice for use of _FillValue and/or
valid_range variable attributes? At one point, I think there was a
performance cost if valid_range was supplied, and I would like to avoid
that, if it's still an issue.
Hi Ellyn,
According to CF Trac Ticket #31 (slated for inclusion in the update to CF 1.7),
the way to cache minimum maximum values in metadata is to use an attribute
named actual_range and store them as a pair.
(I kind of think this is a bad idea, and wish that ticket was still open. I
missed
12 matches
Mail list logo
