Dear Bangyou
Yes, CF has conventions for this. See Section 7.3 on "Cell methods". However,
I'm sorry to say I don't understand your example - I must have misunderstood
something elementary. For instance, if you average over two of the dimensions,
such as sites and sowing, won't there be 3*3=9 values, not 54? Such an average
would be stored in CF as
dimensions:
meansites = 1;
meansowing = 1;
water = 3;
genotype = 3;
float meanyield(meansites, meansowing, water, genotype);
meanyield:cell_methods="meansites: mean meansowing: mean";
However this does not indicate what the meansites and meansowing axes are.
They should describe themselves somehow e.g.
float meansites(meansites);
meansites:long_name="site";
Best wishes
Jonathan
> Our data is structured as follows:
>
> dimensions:
> sites = 3;
> sowing = 3;
> water = 3;
> genotype = 3;
> variables:
> float yield(sites, sowing, water, genotype);
>
> We would like to apply a method (e.g. mean) to any combinations of
> dimensions. For this data set, we will store mean yields for
>
> 1. all 81 yields: *1* value.
> 2. average for three dimensions, sites, sowing, water and genotype: *12
> *values.
> 3. average for two dimensions, sites and sowing, sites and water, sites
> and genotype, sowing and water, sowing and genotype, water and
> genotype: *54 *values
> 4. average for one dimension: *324* values.
>
> Do some conventions already designed? Or do we need to design some
> new conventions? Any suggestions are welcome.
>
> Best regards,
> Bangyou
>
> --
> *Dr. Bangyou ZHENG*
>
> Postdoctoral Research Fellow
> CSIRO Plant Industry
> Queensland Bioscience Precinct
> 306 Carmody Rd
> St Lucia, QLD, 4067, AUSTRALIA
> Tel: +61 (0)7 3214 2620
> Fax: +61 (0)7 3214 2920
> Email: Bangyou.Zheng at csiro.au
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
----- End forwarded message -----
Received on Thu Dec 20 2012 - 01:01:26 GMT
