NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Forgive the ignorance, I'm a relative newbie... I have data for which I guess the Mathtype would be (row,col) -> value, where value is some dependent entity such as temperature or precip (the data is output from the MM5 meterological app). The 'grid' of data points is 126 rows by 150 columns, representing some part of the earth, so 'rows' equate to latitude (??) For my domainSet I have been using just Integer2DSet. Of course in my plot I will get axes labelled 0-126 (Y axis) and 0-150 (X axis). In this state, I cannot locate any true lat/lon value on my grid. Now, I know that the data grid came from a LambertConformalConic projection, and I know all the parameters to build the visad class of that name. So for any (row,col) indices in my data matrix, I can derive the actual lat + lon of that 'grid point'. The grid contains the center of the projection, so if I use a northing and easting I can get my grid origin, ie. row = col = 0, to line up with the correct lat/lon. In the projected grid, the grid points are 36km apart in both X and Y. What I cannot understand is how to use knowledge of the coordinate system to correct the axes and labelling on the plots. Thus, the Y axis would be latitude, scaled between say 30 and 60 degrees, and similar for X/lon. I have so far studied all the code in LambertConformalConic, the MapProjection, the InverseCoordinateSystem, etc and pored through the mail list archive. I'm not sure if I need a display side coordinate system or not. In truth, I am confused about what information I have! Any help very gratefully appreciated. Stuart
visad
archives: