Error when using "pspectrum" with "spectrogram" option.

Hello,

I am trying to plot the pspectrum of a signal, using the following command: pspectrum(SmallY,fs,'spectrogram') and I get the error listed below. SmallY is a 1x100,000 double array (here attached). What puzzles me most is that the same command works for another array, which is a 1x 9000 double (attached here as normal.mat).

Could someone please help with this? I don't see what is the difference between the two arrays, why is one working and the other not. Below, I am also listing the "code", although it's not much. Of course, the same code goes for 'normal' array.

load SmallY

fs=100000;
pspectrum(SmallY,fs,'spectrogram')

-----------------------------------------------------------------------------------------

Error using matlab.graphics.axis.Axes/set

Error setting property 'CLim' of class 'Axes':

Value must be a 1x2 vector of numeric type in which the second element is larger than the

first and may be Inf

Error in signalwavelet.internal.convenienceplot.plotTFR (line 178)

set(ancestor(hndl,'axes'),'CLim',[Pmax plotOpts.threshold]);

Error in pspectrum>displaySpectrogram (line 1078)

signalwavelet.internal.convenienceplot.plotTFR(t,f,10*log10(abs(P)+eps),plotOpts);

Error in pspectrum>computeSpectrogram (line 451)

displaySpectrogram(T,F,P,opts.IsNormalizedFreq,opts.MinThreshold,FRES,TRES);

Error in pspectrum (line 251)

[P,F,TORPWR] = computeSpectrogram(opts, nargout==0);

Error in test_27apr (line 17)

pspectrum(SmallY,fs,'spectrogram')

---------------------------------------------------------------------------------------------------------------

Answer 1

[Deepak Gupta]

Hello Ana,

There is a different between data sets normal and SmallY.

In dataset SmallY, values are extremly small. i.e. max value = 3.0620e-16 and min value = -3.0620e-16. This is what causing the error and pspectrum is unable to support such small numbers.

You can normalise your data before pspectrum to avoid this error.

SmallY = SmallY/max(abs(SmallY));
fs=100000;
pspectrum(SmallY,fs,'spectrogram')

Hope this helps.

cheers.

[Ana Caramete]

Hello,

Thank you very much! Normalising solved the problem indeed.