Fish-Soup

Hi That's fair enough, my team do plan to upgrade to 3 in the future but we have a fair bit of legacy code to update. Cheers anyways

Cheers thanks for that.

Ahh I don't think I quite understood how flexible pipelines where. I'll definitely take a look into this. Apart from the ability to warm start (after the MARS bit) can...

Hi so I've tried to put this into practice and its been a partial success. I have tried Earth=>Ridge, Earth=>Lasso and Earth => Elastic net pipelines. In the example case...

So I've just tried inserting StandardScaler into the pipelines but it doesn't seem to have helped. i think this is the right track though earth_net= Pipeline([('earth',Earth(max_degree=4,enable_pruning=False)),('std_scal',StandardScaler()),('enet',ElasticNet())]) So as i understand...

Cool sorry to take so long to reply. It appears that one of the columns generated in model1.transform(X) is a set of one's. Which obviously can't be normalised. This doesn't...

so It turns out my mistake was assuming that the alpha parameter in elastic net was the same as the one in ridge regression. because of the way it's structured....

@jcrudy I think maybe I'm being unclear or at least I don't understand how changing the loss function would help.... let's say I have some relationship such that d =...

I could do what you suggest. But I'd rather the function was fitted in MARS. for example I know that d has a nonlinear relationship with t* so I wonder...

I'm thinking about a custom loss function. I suppose I have 2 goals. 1) make sure time lagged data is weighted less the longer the lag. 2) I hope by...