ARMAView¶
-
class
ARMAView: public tsa::BaseView¶ ARMA view: container for (vectorial) ARMA processes.
A view for ARMA parametrization. It defines a general (V)ARMA process, which can be written as
\[ \sum_{k=0}^{p} A^{(k)} \vec{y}_{n-k} = \sum_{k=0}^{q} B^{(k)} \vec{x}_{n-k} \]where A,B are square matrix of dimension d equal to the dimension of the input and output vectors x,y. If the order of the part MA q is equal to zero the process is an AR process. If the order of the AR part p is equal to zero the process is an MA process. Note that the matrix \(A^{(0)}\) is assumed to be the identity.Getters
-
const double &
GetAR(int i, unsigned int channel = 0) const¶ This method gives the value of the AR[i] coefficient for one of the channels. It is assumed that the VARMA process is diagonal, which means that there is and independent ARMA process for each channel.
- Return
the value of the AR[i] coefficient
- Parameters
i: the index of the AR coefficientchannel: the channel
-
const double &
GetMA(int i, unsigned int channel = 0) const¶ This method gives the value of the MA[i] coefficient for one of the channels. It is assumed that the VARMA process is diagonal, which means that there is and independent ARMA process for each channel.
- Return
the value of the MA[i] coefficient
- Parameters
i: the index of the MA coefficientchannel: the channel
-
const double &
GetVAR(int i, unsigned int channel1, unsigned int channel2) const¶ This method gives the value of the \(A^{(i)}_{jk}\) coefficient for the VARMA process
- Return
the value of the coefficient \(A^{(i)}_{jk}\)
- Parameters
i: the index of the VAR coefficientchannel1: the first channel (index j)channel2: the se channel (index k)
-
const double &
GetVMA(int i, unsigned int channel1, unsigned int channel2) const¶ This method gives the value of the \(B^{(i)}_{jk}\) coefficient for the VARMA process
- Return
the value of the coefficient \(B^{(i)}_{jk}\)
- Parameters
i: the index of the VMA coefficientchannel1: the first channel (index j)channel2: the se channel (index k)
-
unsigned int
GetArOrder() const¶ Get the order of the AR part of the process.
- Return
the order of the AR part of the process
-
unsigned int
GetMaOrder() const¶ Get the order of the MA part of the process.
- Return
the order of the MA part of the process
-
unsigned int
GetChannels() const¶ Get the dimension of the VARMA process.
- Return
the dimension of the VARMA process
Setters
-
void
SetAR(int i, double v, unsigned int channel = 0)¶ This method sets the value of the AR[i] coefficient for one of the channels. It is assumed that the VARMA process is diagonal, which means that there is and independent ARMA process for each channel.
- Parameters
i: the index of the AR partv: the new value of AR[i]channel: the channel
-
void
SetMA(int i, double v, unsigned int channel = 0)¶ This method sets the value of the MA[i] coefficient for one of the channels. It is assumed that the VARMA process is diagonal, which means that there is and independent ARMA process for each channel.
- Parameters
i: the index of the MA partv: the new value of MA[i]channel: the channel
-
void
SetVAR(int i, double v, unsigned int channel1, unsigned int channel2)¶ This method sets the value of the \(A^{(i)}_{jk}\) coefficient for the VARMA process
- Parameters
i: the index of the VAR coefficientv: the new value of \(A^{(i)}_{jk}\)channel1: the first channel (index j)channel2: the se channel (index k)
-
void
SetVMA(int i, double v, unsigned int channel1, unsigned int channel2)¶ This method sets the value of the \(B^{(i)}_{jk}\) coefficient for the VARMA process
- Parameters
i: the index of the VAR coefficientv: the new value of \(B^{(i)}_{jk}\)channel1: the first channel (index j)channel2: the se channel (index k)
-
void
SetOrder(unsigned int maxP, unsigned int maxQ)¶ This methods sets the order (p,q) of the (V)ARMA process
- Parameters
maxP: the order of the AR part of the processmaxQ: the order of the MA part of the process
-
void
SetChannels(unsigned int channels)¶ This method sets the dimension of the VARMA process
- Parameters
channels: the dimension of the VARMA process
-
const double &