Variogram.fit() - Setting parameters

[Kim Lewandowski]

Hi Chris,

When using variogram.fit(), I realise that I have the option to change parameters before I 'save to guess' however, I am unsure whether I need to change the parameters 

Furthermore, when I do change the parameters I am unable to get a fit that works for both short and long-lag. It is either a good fit for short but a bad fit for long or vice versa.

Can you please advise how to best do this to ensure my parameters are set correctly?

Thanks

Kim

[Christen Fleming]

Hi Kim,

1. variogram.fit() is deprecated in favor of ctmm.guess(), but it likely won't make any difference here.
2. You generally don't have to fiddle with the parameters guesses. It only needs to be in the ballpark.
   
3. You should ultimately be using the fitted parameters. If they don't reflect the empirical variogram (within uncertainty), then there could be different causes for that. Feel free to post the output plots here. A big culprit is often non-stationarity and, in particular, dispersal behavior.

Best,

Chris