You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
'nan' is possibly due to the exceed of Python's largest number limit. When the trakectory length T is long + both f() and h() are highly non-linear + the noise variances Q and R are high, the setting becomes very difficult for UKF to accurately predict. In this case, if we calculate square loss accumulated over the long trajectory, it may easily become a very large number.
Dear Authors,
I have a question about it. Why UKF‘s results are "nan" in the Loreanz attractor with Noisy non-linear observations ?
Looking forward to your reply. Thank you!
Best wishes!
The text was updated successfully, but these errors were encountered: