This action writes out information about the optimization to a trace file every skip steps, starting with the state just before the optimization (step 0). The type of information depends on the type of optimization being carried out, but generally includes the iteration number, energy values, and atomic shifts.
output can be a standard Python file object, to which the trace is written, or a file name. In the latter case, a file with that name is created (overwriting any existing file). If output is not specified, the trace is written to the logfile instead.
Example: See conjugate_gradients() command.