Closed foufa89 closed 5 years ago
Hi @foufa89, could you provide more context to the problem? Did you generate both HV and EPS results for your experiments?
I think this from the README is relevant to the problem:
All the quality indicators are computed in this stage except for the hypervolume (hv) indicator. If you want to include it in the processing, you have to execute step 6 of MATLAB Experiment procedure (above).
Hi @ash-aldujaili
Thanks for your prompt reply,
Yes, HV results are generated and the run_postprocessing
works for indicators = ['eps', 'igd', 'hv']
This problem appears with GD indicator, separable functions (I think)
Hey @foufa89 is the issue fixed?
Hey @ash-aldujaili, Ramadan Mubarak;
Not really, Now the post-processing is completed, but the resulting GD-plots are stable horizontal lines.
Thank you.
It is possible the algorithm is not hitting any of the GD targets. Consequently, its profile will be flat. Try giving the algorithm more function evaluations.
Ok, I will try this.
Hi,
I Got this error when executing
run_postprocessing.py
.left==rigth
, is it a precision problem in this part offancyplot_bigFont.py
?I think, probably,
np.arange
works on a small scaleHow we can fix it? PS: for the epsilon, IGD, and HV indicators we don't have this error message, as it's shown on the figure it appears with GD indicator