openclimatefix / uk-pv-national-xg

National PV forecasting using Gradient Boosted Methods.

4 stars 3 forks source link

Change scaling to relative #95

Closed jacobbieker closed 1 year ago

jacobbieker commented 1 year ago

Fix #94

Pull Request

Description

Fixes #94

How Has This Been Tested?

Please describe the tests that you ran to verify your changes. Provide instructions so we can reproduce. Please also list any relevant details for your test configuration

[ ] Yes

Checklist:

[ ] My code follows OCF's coding style guidelines
[ ] I have performed a self-review of my own code
[ ] I have made corresponding changes to the documentation
[ ] I have added tests that prove my fix is effective or that my feature works
[ ] I have checked my code and corrected any misspellings

peterdudfield commented 1 year ago

Should the lower scaling value be changed too?

jacobbieker commented 1 year ago

Should the lower scaling value be changed too?

As in, be higher than 0.4?

peterdudfield commented 1 year ago

Should the lower scaling value be changed too?

As in, be higher than 0.4?

Becasue you are now timings by 0.4 the delta, should it be >1, so that the delta is bigger. This would then make the lower plevel lower than it was originally, which I think is the effect we wanted?

jacobbieker commented 1 year ago

Should the lower scaling value be changed too?

As in, be higher than 0.4?

Becasue you are now timings by 0.4 the delta, should it be >1, so that the delta is bigger. This would then make the lower plevel lower than it was originally, which I think is the effect we wanted?

Oh yeah, probably then make it the inverse? So 2.5?

peterdudfield commented 1 year ago

Should the lower scaling value be changed too?

As in, be higher than 0.4?

Becasue you are now timings by 0.4 the delta, should it be >1, so that the delta is bigger. This would then make the lower plevel lower than it was originally, which I think is the effect we wanted?

Oh yeah, probably then make it the inverse? So 2.5?

Just looking at today's results, I would go for something like 1.6, keep it the same for the moment.

I imagine we might have to adjust these as we get some live feedback. From todays values, they didnt look liek the needed adjusting at all

jacobbieker commented 1 year ago

Should the lower scaling value be changed too?

As in, be higher than 0.4?

Becasue you are now timings by 0.4 the delta, should it be >1, so that the delta is bigger. This would then make the lower plevel lower than it was originally, which I think is the effect we wanted?

Oh yeah, probably then make it the inverse? So 2.5?

Just looking at today's results, I would go for something like 1.6, keep it the same for the moment.

I imagine we might have to adjust these as we get some live feedback. From todays values, they didnt look liek the needed adjusting at all

Okay, sounds good, I'll change that then to that.