Ethan Lew

@nsde I have not forked this as my use case for it didn't work out. It should be pretty straightforward to load up the model in the old python and...

Comparisons | | Board Size [s] | Solver | Previous Time [s] | New Time [s] | Multiplier | |---:|-----------------:|:---------|--------------------:|---------------:|-------------:| | 0 | 100 | pulp | 0.2395 | 0.1014...

Same here

Update: for me, this was solved just by updating my environment.yml (environment description) to something that can be resolved by python 3.12 on ubuntu 22.04. I did not need to...

@stanleybak , that's a great feature. We've seen the need to normalize trajectories in the past for library learning over Legendre polynomial candidate functions since the functions are only orthogonal...

I guess #59 doesn't normalize the evaluation but only the data for the estimator `Estimator.fit` method.

@TUMcps Idk what reset means. We currently run a simulation between each step and hold the inputs constant (ZOH). Would the idea be to run the simulation at a finer...

Ah yes, this is coming back to me. Yes, we reset "all" in continuous mode because we pass it to an ODE solver that evaluates $\dot{\mathbf x} = f(\mathbf x,...

I fixed `n_obs` and `rank`. I don't think it makes sense to have lengthscale in the form of (start, stop, step) because it's a continuous hyperparameter (just (start, stop)) and...

@Abdu-Hekal This is an interesting point. I agree for KIC with constant inputs you should not need that last state. However, there are formulations that would need the extra input,...