Fixing the note from submission 5. Large Data Input and Output

"Finally, in accordance with convention, we calculate the latitude and longitude values by converting metres to degrees and write them to our netCDF file if we are using degrees as our unit:"

[x] It is interesting to have spherical output but this was not required.

"First, the variable to store the ID is created on the name that the netCdf uses as an alias for the longitude. In this case, the aliases are lon, longitude, x, X and a check is made to see if any of them are included in the file."

[x] The input was given in Cartesian coordinates, i.e., in meters. Otherwise the values wouldn't make much sense either:

[alex@jenalex tmp]$ ncdump artificialtsunami_bathymetry_1000.nc  | head -n 500
netcdf artificialtsunami_bathymetry_1000 {
dimensions:
    x = 1000 ;
    y = 1000 ;
variables:
    float x(x) ;
    float y(y) ;
    float z(y, x) ;

// global attributes:
        :Conventions = "COARDS" ;
data:

 x = -4995, -4985, -4975, -4965, -4955, -4945, -4935, -4925, -4915, -4905, 
    -4895, -4885, -4875, -4865, -4855, -4845, -4835, -4825, -4815, -4805, 
[...]

        case VarType::CHAR:
            outData[i].array = new char[length];
            break;
        case VarType::SHORT:
            outData[i].array = new short[length];
            break;
        case VarType::INT:
            outData[i].array = new int[length];
            break;
[...]

[x] This is a bit over-engineered for the purpose of tsunami simulations.

(tsunami_lab) tsunami_19@LeChuck:~/05_large_io/gerlach_hofer/Tsunami-Simulation/build$ ./sanitize_test 
 writing to 'solutions/simulation/WriteNetCDF.test.nc'
 writing to 'solutions/simulation/WriteNetCDF.test.nc'
 writing to 'solutions/simulation/WriteNetCDF.test.nc'
 writing to 'solutions/simulation/WriteNetCDF.test.nc'
 writing to 'solutions/simulation/WriteNetCDF.test.nc'
 [...]

[x] unit tests should be silent.

" -M use meters as unit for the X-Axis and Y-Axis instead of the Longitude and Latitude units degrees_east and degrees_north."

[x] Degrees would require you to do the mapping to Cartesian coordinates internally and to numerically work on curved mesh.

    l_setup = new tsunami_lab::setups::TsunamiEvent2d( "resources/artificialtsunami_bathymetry_1000.nc",
                                                       variables,
                                                       "resources/artificialtsunami_displ_1000.nc",
                                                       variables,
                                                       l_scaleX,
                                                       l_scaleY );

[x] Hardcoding the file paths is not too elegant. From the documentation of your code its not clear what is being computed..

#include <cmath>
#include <math.h>

[x] Why do we require both includes?

RivinHD / Tsunami-Simulation

Fixing the note from submission 5. Large Data Input and Output #53