For Gaussian to print SCF convergence information, it is necessary to specify additional print, by using #P in the route section.

Using GaussSum, open the file you want to examine. Usually this will be the output file of a job which is currently running.

Choose SCF.py and Click on the GaussSum logo to run the script. A graph will be plotted of the progress of the SCF convergence versus step number. Details of the SCF cycles are written to the screen. Where there is information on several SCF convergence cycles in a log file (e.g. in the case of a geometry optimisation), GaussSum plots the one closest to the end of the file.

The progress of the SCF convergence is calculated by considering the convergence criteria. For example, for Gaussian, these criteria are the RMS density, MAX density and change in energy. Each of the current values is divided by the target values. The log is taken of any results greater than 1. The 'progress value' is then calculated by summing the logs. Note that convergence is achieved when the progress value equals 0.

You may wish to leave out the initial points, so as to focus on the last few points. You can do this by entering a non-zero value into the box labeled "Leave out the first n points", and re-running the script.

Note that, at the moment, this procedure does not work for SCF=QC (Gaussian).