Re: annual mean and contour

Excuse me. It was pointed out that I am wrong. The ispan call does indeed return 85 values. However, in spite of the fact that you have nlat set to 85, there is nothing in the code that shows what is the actual dimensionality of the 'var' variable.  You need a printVarSummary(var) to verify that you have what you think you have for its shape.  -dave On Feb 7, 2012, at 2:04 PM, David Brown wrote: Well the error message is telling you what is wrong: ispan(-42,42,1) produces 84 values, whereas your nlat value is 85.  However, you should put some 'printVarSummary' messages in your code to help show you what is the problem.  -dave On Feb 7, 2012, at 1:45 PM, Xiaoyan Ma wrote: Hi I want to calculate the annual mean from 12 months hdf files, and then make a contour, below is what I did. There is a fatal error as below --------------------------------------------------------------------------------------------- fatal:Coordinate variables must be the same dimension as their dimension fatal:Execute: Error occurred at or near line 43 in file ------------------------------------------------------------------------------------------- line 43 is "   var&lat        = ispan(-42,42,1)*dlat " Also the values of var is not correct. can anyone help to fix the problem? Thanks, Xiaoyan ================================    nlat  = 85    nlon  = 72    ntim = 12    do nt=0,ntim-1    fname = m+sprinti("%0.2i", nt+1)+".hdf"    print(fname)    in = addfiles(fname, "r") ;; load all the files    var0 = in[:]->variable_name    end do    var = dim_avg_n(var0, 0)    dlat  = 170./nlat    dlon  = 360./nlon    var!0          = "lat"    var&lat        = ispan(-42,42,1)*dlat    var&lat@units  = "degrees_north"    var!1          = "lon"    var&lon        = ispan(-36,35,1)*dlon    var&lon@units  = "degrees_east" ... _______________________________________________ ncl-talk mailing list List instructions, subscriber options, unsubscribe: http://mailman.ucar.edu/mailman/listinfo/ncl-talk _______________________________________________ ncl-talk mailing list List instructions, subscriber options, unsubscribe: http://mailman.ucar.edu/mailman/listinfo/ncl-talk