While working with bits of the code for a small thing by myself, I found, as I suspected, that it is necessary to have bounds on the parameters for the Confined and Hop models. Once these are implemented, the fits converge very nicely. Only issue, they have to be based on the Brownian model fit. So if the Brownian model gives a D_B and a delta_B, the bounds for the confined model should be:
D_micro: between 0 and 1.5*D_B
delta: between 0.75*delta_B and 1.25*delta_B
tau: between 0 and np.inf
hop diffusion:
D_Macro: between 0 and 1.5*D_B
D_micro: between 0 and 1.5*D_B
delta: between 0.75*delta_B and 1.25*delta_B
tau: between 0 and np.inf
Once this issue and #32 are solved, I think we are pretty much good to go.
While working with bits of the code for a small thing by myself, I found, as I suspected, that it is necessary to have bounds on the parameters for the Confined and Hop models. Once these are implemented, the fits converge very nicely. Only issue, they have to be based on the Brownian model fit. So if the Brownian model gives a D_B and a delta_B, the bounds for the confined model should be:
D_micro: between 0 and 1.5*D_B delta: between 0.75*delta_B and 1.25*delta_B tau: between 0 and np.inf
hop diffusion: D_Macro: between 0 and 1.5*D_B D_micro: between 0 and 1.5*D_B delta: between 0.75*delta_B and 1.25*delta_B tau: between 0 and np.inf
Once this issue and #32 are solved, I think we are pretty much good to go.