Above is output from SAS.
What would be the corresponding ARIMAX equation? I would appreciate if someone could help me write the mathematical equation, preferably in the following form:
$$
Y(t)= ay(t-1) + by(t-2) + ldots + z
$$
where $a, b, c$ are the coefficients, and $z$ is any error term.
Best Answer
From what I know of the ARIMA model (please correct me anyone if I'm wrong) is that it is equivivalent to "pre-filter" (differencing) the data $y(t)$ and set $y_{new}(t) = y(t)-y(t-1)$. Then fit an ARMA model to this new data. This would mean your ARMA model:
begin{equation} y_{new}(t) = (1-q^{-1}) y(t) = frac{C}{A}e(t), end{equation} where $q^{-1}$ is the difference operator $q^{-1}y(t) = y(t-1)$ and $e(t)$ is noise.
This means you get the model (in $y$) as begin{equation} (1 – q^{-1})(1+a_1q^{-1}+ldots+a_{n_a} q^{-n_a})y(t) = (1 + c_1 q^{-1} + ldots + c_{n_c} q^{-n_c})e(t). end{equation}
Similar Posts:
- Solved – Simple linear model with autocorrelated errors in R
- Solved – How to fit a negative binomial distribution in R while incorporating censoring
- Solved – How to fit a negative binomial distribution in R while incorporating censoring
- Solved – How to calculate greater than or equal to in R using ppois?
- Solved – C# ARMA library