This software provides different parametric estimation methods for stochastic differential equations (SDE), namely the Vasicek and CKLS models, as well as functions to simulate paths of the SDEs. The methods included are:
- Exact maximum likelihood (EML)
- Euler method (DML)
- Local linearization (LL)
- Hermite polynomial expansion (HP)
- Generalized Method of Moments (GMM)
- Kalman Filter (KF)
- Markov Chain Monte Carlo (MCMC)
Get the package from GitHub:
# Install the package
library(devtools)
install_github("alejandralopezperez/estsde")
# Load package
library(estsde)
The following is an example of data generation and estimation:
# Generate data from a CKLS model
set.seed(789)
x <- rCKLS(480, 1/12, 0.09, 0.08, 0.9, 1.2, 1.5)
# Parameter estimates
est.CKLS.KF(x)
#>
#> Call:
#> dX_t = (alpha - kappa X_t)dt + sigma X_t^gamma dW_t;
#>
#> Coefficients:
#> Estimate Std. Error
#> alpha 0.0795 0.0192
#> kappa 0.8979 0.2530
#> sigma 1.3393 0.3890
#> gamma 1.5496 0.1176