3
$\begingroup$

May I know what $x,y,z$ in ARProcess[x,{y},z] mean?

ma = EstimatedProcess[detrend, ARProcess[1]]
(* Out: ARProcess[-1.14364*10^-13, {0.235799}, 0.00209441] *)
$\endgroup$

2 Answers 2

3
$\begingroup$

Did you check the manual for ARProcess?

You have a numbers (not matrices), thus you need this and this rules:

ARProcess[{a1,…,ap},v] represents a weakly stationary autoregressive process of order p with normal white noise variance v.

ARProcess[c,…] represents an AR process with a constant c.

Thus, in your case you have an AR process of degree 1 with constant c equal to -1.14364*10^-13 (zero means), coefficient of an AR process 0.235799 and a variance of residuals equal to 0.00209441.

$\endgroup$
2
$\begingroup$
Manipulate[
 Module[{f = RandomFunction[ARProcess[x, {y}, z], {1, 10^2}]},
  ListPlot[SeedRandom[1234]; f,
      Filling -> Axis, 
      PlotRange -> {Automatic, {-0.5, 0.5}}]],
 {{x, -a}, -1, 1},
 {{y, b}, -3*b, 3*b},
 {{z, c}, $MinMachineNumber, 20*c}, 
 Initialization :> (a = 1.14364*10^-13; b = 0.235799; c = 0.00209441)]

enter image description here

ARProcess[x, {y}, z] represents a first order autoregressive process. If you play around with the RandomFunction defined by the process in the code above, you'll see that $x$ is a constant offset, $y$ is a process parameter, and $z$ represents the variance.

$\endgroup$
1
  • $\begingroup$ may I know the name of those? $\endgroup$ Commented Oct 12, 2015 at 14:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.