Hi all, There have been a couple of questions lately about working with
attributes in datasets. Here's a note about some of the commands
and options you may not know about. All of this is here in the
documentation: http://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/commands-reference/SET#_VPID_474
SET ATTRIBUTE/LIKE= This command takes all of the
attributes from var1 and copies them to attributes belonging to
var2, so if you're defining a new variable based on a file
variable, you'll be able to keep the units, title, and anything
else that the file variable has. (It doesn't inherit
missing-value flags - those are defined as always when var2 is
defined.) yes? set attribute/like=fvar var2 (See also LET/LIKE= which applies the same setting when defining variables.)
SET ATTRIBUTE/OUTPUT This marks attributes for writing
out to netCDF files. By default Ferret/PyFerret only write the
title, units and history, but this lets you set properties so they
are written out. To set a property so that one attribute or all
attributes on a variable will be written out: yes? set attribute/output fvar.comment (See also DEFINE ATTRIBUTE/OUTPUT for setting the output property when defining a new attribute)
To work with global attributes, use a dot. This acts as a
virtual "global-attribute" variable name. Global attributes are
written when first creating a file, writing variables that belong
to that file (writing user-variables based on file variables will
also trigger writing the global attributes from that file.) To
write a couple of attributes: yes? set attribute/output ..investigators Or set them all to be written out yes? set attribute/output . Finally, to add new global attributes to an existing dataset, uses SAVE/ASGLOBAL which was new in version 7.3
Happy Ferreting, Ansley |