The graphical interface has long offered some generic methods for computing one, some, or all Nash equilibria, where available.
Similarly, for many types of problem, scipy offers generic routines which hide the selection of, e.g. a specific algorithm. See, for example, function minimisation or ODE solving.
It would be useful to offer something similar in pygambit.
In doing so it will be worth giving some thought as to how best to keep this in sync between the graphical interface and pygambit. At the moment the graphical interface makes calls out to the command-line tools, but perhaps this should be reviewed - which would allow for having the logic for EC methods to be maintained in a single location.
The graphical interface has long offered some generic methods for computing one, some, or all Nash equilibria, where available.
Similarly, for many types of problem,
scipyoffers generic routines which hide the selection of, e.g. a specific algorithm. See, for example, function minimisation or ODE solving.It would be useful to offer something similar in
pygambit.In doing so it will be worth giving some thought as to how best to keep this in sync between the graphical interface and
pygambit. At the moment the graphical interface makes calls out to the command-line tools, but perhaps this should be reviewed - which would allow for having the logic for EC methods to be maintained in a single location.