[vtkusers] Re: writing several scalars in one file (py and C)...

[fred]

Dominik Szczerba a écrit :
> InsertNextValue() reallocates memory if needed, probabely logarithmically. If 
> you are not fine with it, and know the size beforehands, just set the size of 
> the array at the beginning and write the buffer with [] operator instead of 
> appending/reallocation. Read the vtkArray documentation for details.
> Dominik
>
>   
Ok.

Does this few lines look right ?

def write_vtkdata_scalars(writer, data, format, name, Nx, Ny, Nz):
    print name,
    vtkdata = dico_vtkArray_fmt[format]()
    vtkdata.SetNumberOfTuples(Nx*Ny*Nz)
    vtkdata.SetName(name)
#     vtkdata.SetNumberOfComponents(1) # set by default
    for k in range(Nz):
        for j in range(Ny):
            for i in range(Nx):
                vtkdata.SetTuple1(k*(Nx*Ny)+j*Nx+i, data[i][j][k])
    if (writer.GetPointData().GetNumberOfArrays() == 0):
        writer.GetPointData().SetScalars(vtkdata)
    else:
        writer.GetPointData().AddArray(vtkdata)

I use write_vtkdata_scalars() as

    write_vtkdata_scalars(writer, data[0], format, 'raw', Nx, Ny, Nz)
    write_vtkdata_scalars(writer, data[1], format, 'filtered', Nx, Ny, Nz)
    write_vtkdata_scalars(writer, data[2], format, 'residue', Nx, Ny, Nz)

Can I do this in a better way ?

Thanks in advance.

Cheers,

-- 
/F