maximize
  logexp_sum(matrix_allscenarios)
Solver: precision = 9
Value:
  pr_pen(0,L(matrix_scen_with_label1)-L(matrix_scen_with_label0))