Details of each datatype used by the Dosimetry App can be found at the Dosimetry Manifest Guide.

Some more complex data types have detailed descriptions and example usages shown below:

A Dij matrix is a container used to store the dose from each beamlet to each calculation point (Note: for SOBP beams, a beamlet is simply a small square area of the field; whereas for PBS beams, a beamlet is a physical pencil beam). It's stored sparsely for efficiency purposes as each beamlet will only provide dose to a small subset of the field points. Since the matrix is stored sparsely, it's important to understand the proper means for extracting the data it contains. The matrix data is stored as a list of dij_entry values. These values are properly indexed to beamlets and points by using the dij_row values also held as a list within the dij_matrix container. There is one dij_row value for each dose point and each value contains the offset position for accessing the list of entries as well as the number of entries for this point. This is best explained by reviewing the code example provided below that shows how to loop over the complete dij_matrix, summing the dose to compute the total dose to each calculation point.

std::vector<double> pdose_total(dij.n_points, 0.0);
for (uint32 i = 0; i != dij.n_points; ++i)
{
    dij_row const& row = dij.rows[i];
    dij_entry const* entry = dij.entries.elements + row.offset;
    dij_entry const* end = entry + row.n_entries;
    // Loop over the beamlets (entries) for point i
    for (; entry != end; ++entry)
    {
        // dose to this point from entry->beamlet_index is entry->dose, so add it's value to the point dose
        dose_total[i] += entry->dose;
    }
}