rsteca / sklearn-deap

Use evolutionary algorithms instead of gridsearch in scikit-learn
MIT License
770 stars 130 forks source link

Issue27 [Untested] #29

Closed ryanpeach closed 7 years ago

ryanpeach commented 7 years ago

Try this to get the first evaluation, also added the selection to capture all available data.

Is someone willing to test this?

ryanpeach commented 7 years ago

I did this from the web, I don't have access to a test environment at the moment and wont have time for a while to run it, so I'll just leave this here since it's pretty simple now that I think about it.

ClimbsRocks commented 7 years ago

Thanks for this! i'll try testing it out later today.

rsteca commented 7 years ago

I tried it out but it doesn't seem to work for me. I get the following error:

AttributeError: 'numpy.float64' object has no attribute 'history_index'

ryanpeach commented 7 years ago

Ok I'll check it out now.

ryanpeach commented 7 years ago

So you're right, you can't use the evaluation history decorator. But I have updated the notebook to make things a little more clear :)