## NARMA modeling

Purpose

Description

Macro Synopsis

Modules

Related Functions

References

### Purpose

Estimate the coefficients of a NARMA model

### Description

"NARMA modeling" computes the least-squares estimates of the
coefficients of a Nonlinear Auto-Regressive Moving Average (NARMA)
model involving linear and bilinear terms of lagged values of the
response signal,

The first input signal (y) to "NAR modeling" is regarded as
response, the second one (x) is the stimulus. The output signal r is
the residual, i.e. the estimation error.
Parameters are:

- the linear order K of the response process
- the bilinear order M of the response process and
- the linear order L of the stimulus process

Besides the estimated coefficients, a residual constant term, the mean of the
residues, the residual variance, and the normalized residual variance
are displayed in the message window.
**Warning:** *Note that if the input signals have different x-axis scales (sampling periods), the signal with the largest scale will be adapted automatically by interpolating between successive sample points. The type of interpolation can be set in the Basic Options menu.*

### Macro Synopsis

`r = NARMAmodel([y,x],K,M,L);`

signal r,x,y;

int K,L,M;

### Modules

Statistics

### Related Functions

AR simulation, ARIMA modeling,
ARIMAX modeling, ARMA model order,
ARMA modeling, ARMA simulation,
ARMA spectral density, ARMAX modeling,
ARMAX simulation, NAR modeling.

### References

Ljung [10], Conover [11], Graybill
[12], Hollander/Wolfe [13]