Calculate the objective (i.e. the negative log-likelihood)
function using a wrapper for a given stochastic process
| Parameters: |
-
params
(list)
–
The input values representing parameters for the stochastic process.
-
process
(Any | None, default:
None
)
–
An instance of a stochastic process object
-
observations
(DataFrame | DataFrame | None, default:
None
)
–
The observation data used for calculating the log-likelihood
-
delta
(float, default:
1.0
)
–
The sampling frequency or time step used in the observations
|
| Returns: |
-
float
–
the negative log-likelihood value
|
Notes
- If the provided parameters lead to an invalid computation (e.g., due to invalid input or model constraints),
the function returns positive infinity (np.inf).
- The log-likelihood value returned is negated to align with optimization conventions,
where the objective is typically minimized.