[Thread Prev][Thread Next][Index]

Re: [ferret_users] Lat/long to openDAP indices conversion



Hi Tiho,
Ferret makes calls to the NetCDF library for all OPeNDAP I/O requests. Current versions of Ferret are linked with NetCDF v4.1.1. When you open the dataset URL with a USE statement, Ferret reads and stores the dimension and coordinate information as well as the variable names and their attributes. When you request a subset, the conversion between coordinates and indices is done using the coordinate values (which are stored by Ferret as double precision values). Then when you read a subset of data, the appropriate call is made to the NetCDF library to actually read that subset of the data.

-Ansley

On 7/21/2010 9:19 PM, Tiho.Beretovac@xxxxxxxx wrote:
Hi there,

New to ferret, but have tried for a while and have been unable to find an answer to this question:

How does ferret convert the lat long bounding box constraints it uses into indices of individual OpenDAP requests when it fetches the data from an OpenDAP source.

For instance, if you point ferret to an OpenDAP source and ask for a FILL\x=130e:140e (ie. lat long based request), this is translated into an OpenDAP fetch with something like ?mydata[0][32:85][0:499] (ie. the indices that represent the data within the eastings and northings specified in the ferret request).

Are the specifics of this conversion explained anywhere?
Any insight appreciated.

Cheers,

Tiho



[Thread Prev][Thread Next][Index]

Contact Us
Dept of Commerce / NOAA / OAR / PMEL / TMAP

Privacy Policy | Disclaimer | Accessibility Statement