spectrum.eigenvector (Signal Processing Toolbox)

Eigenvector spectrum

Syntax

Hs = spectrum.eigenvector
Hs = spectrum.eigenvector(NSinusoids)
Hs = spectrum.eigenvector(NSinusoids,SegmentLength)
Hs = spectrum.eigenvector(...,OverlapPercent)
Hs = spectrum.eigenvector(...,WindowName)
Hs = spectrum.eigenvector(...,SubspaceThreshold)
Hs = spectrum.eigenvector(...,FFTLength)
Hs = spectrum.eigenvector(...,InputType)

Description

Hs = spectrum.eigenvector returns a default eigenvector spectrum object, Hs, that defines the parameters for an eigenanalysis spectral estimation method. This object uses the following default values:

Property Name
Default Value
Description

NSinusoids
2
Number of complex sinusoids

SegmentLength
4
Segment length

OverlapPercent
50
Percent overlap between segments

WindowName
'Rectangular'
Window name string or 'User Defined' (see window for valid window names). For more information on each window, refer to its reference page.

This argument can also be a cell array containing the window name string or 'User Defined' and, if used for the particular window, an optional parameter value. The syntax is {wname,wparam}.

You can use set to change the value of the additional parameter or to define the MATLAB expression and parameters for a user-defined window (see spectrum for information on using set).

SubspaceThreshold
0
Threshold is the cutoff for signal and noise separation. The threshold is multiplied by lambda _min , the smallest estimated eigenvalue of the signal's correlation matrix. Eigenvalues below the threshold ( lambda _min*threshold) are assigned to the noise subspace.

FFTlength
'NextPow2'
String defining how the number of FFT points is determined. The default is the next power of 2 that is greater than the input length. Other valid values are 'InputLength' and 'UserDefined'. InputLength uses the length of the input signal as the FFT length. UserDefined uses the value provided via the pseudospectrum method or a dspopts object. See dspopts for an example.

InputType
'Vector'
Type of input that will be used with this spectrum object. Valid values are 'Vector', 'DataMatrix' and 'CorrelationMatrix'.

Property Name	Default Value	Description
NSinusoids	`2`	Number of complex sinusoids
SegmentLength	`4`	Segment length
OverlapPercent	`50`	Percent overlap between segments
WindowName	'Rectangular'	Window name string or `'User Defined'` (see `window` for valid window names). For more information on each window, refer to its reference page. This argument can also be a cell array containing the window name string or `'User Defined'` and, if used for the particular window, an optional parameter value. The syntax is `{wname,wparam}`. You can use `set` to change the value of the additional parameter or to define the MATLAB expression and parameters for a user-defined window (see `spectrum` for information on using `set`).
SubspaceThreshold	`0`	Threshold is the cutoff for signal and noise separation. The threshold is multiplied by _min , the smallest estimated eigenvalue of the signal's correlation matrix. Eigenvalues below the threshold (_min`*threshold)` are assigned to the noise subspace.
`FFTlength`	'NextPow2'	String defining how the number of FFT points is determined. The default is the next power of 2 that is greater than the input length. Other valid values are `'InputLength'` and `'UserDefined'`. `InputLength` uses the length of the input signal as the FFT length. `UserDefined` uses the value provided via the `pseudospectrum` method or a `dspopts` object. See `dspopts` for an example.
InputType	'Vector'	Type of input that will be used with this spectrum object. Valid values are `'Vector'`, `'DataMatrix'` and `'CorrelationMatrix'.`

Hs = spectrum.eigenvector(NSinusoids) returns a spectrum object, Hs, with the specified number of sinusoids and default values for all other properties. Refer to the table above for default values.

Hs = spectrum.eigenvector(NSinusoids,SegmentLength) returns a spectrum object, Hs, with the specified segment length.

Hs = spectrum.eigenvector(...,OverlapPercent) returns a spectrum object, Hs, with the specified overlap between segments.

Hs = spectrum.eigenvector(...,WindowName) returns a spectrum object, Hs, with the specified window.

Note Window names must be enclosed in single quotes, such as spectrum.eigenvector(3,32,50,'chebyshev') or spectrum.eigenvector(3,32,50,{'chebyshev',60})

Hs = spectrum.eigenvector(...,SubspaceThreshold) returns a spectrum object, Hs, with the specified subspace threshold.

Hs = spectrum.eigenvector(...,FFTLength) returns a spectrum object, Hs, with the specified way of the determing the FFT length.

Hs = spectrum.eigenvector(...,InputType) returns a spectrum object, Hs, with the specified input type.

Note See peig for more information on the eigenanalysis algorithm.

Examples

Define a complex signal with three sinusoids, add noise, and view its pseudospectrum using eigenanalysis. Set the FFT length to 128.

randn('state',1);
n=0:99;
s=exp(i*pi/2*n)+2*exp(i*pi/4*n)+exp(i*pi/3*n)+randn(1,100);
Hs=spectrum.eigenvector(3,32,95,'rectangular',5);
pseudospectrum(Hs,s,'NFFT',128)

spectrum.cov spectrum.mcov