fix for fitness curve only ever getting updated at the end #44
Conversation
There was a problem hiding this comment.
Hi @davideasaf,
I understand what you are trying to do here. However, I think you have implemented the change at the wrong point in the code. With a slight change, this should be able to be merged.
Regards,
Genevieve.
There was a problem hiding this comment.
I kind of stuffed up this review because I initially misread your changes to the code (and my original code). Sorry about that. Anyway, I understand what you are doing here, but I think you have implemented it in the wrong location. If you could make a slight change to your code, then I will be happy to merge it with the code base.
|
|
||
| if curve: | ||
| fitness_curve.append(problem.get_fitness()) | ||
|
|
||
| # If best neighbor is an improvement, move to that state | ||
| if next_fitness > problem.get_fitness(): |
There was a problem hiding this comment.
@gkhayes , When you say "wrong location" are you referring to the fact that it"s happening before this line? If so, that makes sense.
I believe we both agree on that fact that it should be within the while loop, correct?
There was a problem hiding this comment.
Sorry, I should have been clearer. Anyway, what you have done is in agreement with my thinking.
Instead of updating the fitness_curve before we set the state, we set the fitness_curve after we set the state.
|
I think I addressed the issue. Let me know if that"s right. |
There was a problem hiding this comment.
Hi @davideasaf,
Thanks for making the requested change. It looks good now, so I"m going to merge your PR with the code base and it will be included in the next mlrose release on PyPi.
Regards,
Genevieve.
Currently, the hill climber only ever responds with 1 item in the curve array because the fitness_curve is only ever getting updated at the end of the while loop.