Non-Pareto optimal points plotted when using multiobjective optimisation and gaplotpareto

I'm trying to plot a Pareto curve in a multiobjective optimisation (using gaplotpareto) but the resultant curve has some plot points that are non-pareto optimal (see image). Is this a bug or why are non-Pareto optimal plot points included?

The only options I'm setting are the plot options and the only constraints are lower and upper bounds for the solution variables.

opts = gaoptimset('PlotFcns',{@gaplotpareto,@gaplotscorediversity});

[optimal_sol, optimal_fitness, exitFlag, Output] = gamultiobj(fitnessFun, num_vars,[],[],[],[], lower_bounds, upper_bounds, opts);

[Rakesh Kumar]

This looks like a bug. Can you post the options you are using for this problem? Also, what constraints does your problem have?

Please send a reproduction step to our tech support department for investigation.

[ss rr]

I've edited the question to add the options and constraints I'm using.

Thanks, Sean

[Rakesh Kumar]

If there were non-linear constraints, the solver may accept some dominated points if they are feasible. However, in your case, the entire population should be feasible so this is a bug. I have not seen this before so I have to look at the function or at least the bounds to be able to debug this. Can you send this info to my email?

Thanks, Rakesh

[ss rr]

Yes, if it should only plot the Pareto optimal points then it must be a bug since no matter what the fitness function is, dominated solutions should never make it into the Pareto front set. I'll see if I can recreate it myself first, the curve was generated after several hours leaving it running. I'm running it again now, looks okay so far...