Find Peak Value

Help you find peak value from a lot of data.

Description

It's a simple function. Can help you find peak value from a lot of datas. It contains some demos.

It's my first time to submit file, thanks for your using.

• MATLAB release: MATLAB 6.5 (R13)

[gu wan]

very good

[Jiao Shibo]

You are Genius!

[ma alex]

GOOD!

[Barney de Villiers]

Really Good!
Thanks, It helps me alot!

[Keven Chang]

Great Job.

[Ertu sariaslan]

Thanks for solving my Problem !!!
But I dont understand the sensitivity, can anybody help me ???

[James McDonald]

It worked perfectly for my example. Thanks!!

[Peng Wang]

I just need a function to find peaks. Thank you so much to solve my problem and save my time. It works great. The sensitivity is actually the half-width of the peak. A point is defined as a peak if its value is biger than that of any s points ahead and s points behind.

[Tom Wang]

Not sure if I'm using this correctly but I'm not getting accurate peaks at all. My call to the function is fpeak(1:129, data, 2). "data" has 129 values. I'm getting values of 4.0 0 next to 5.0 1, and 3.0 2. Any suggestions would be appreciated.

[Geng Jun]

Hi,please contact me using this email address!

[ling lin]

It saved my time.

[Alexandre Chirokov]

I added small code (see below) to the getPeak function, now it works even better on my data.
 
if abs(y(i)-avg)<delta
    p=[nan,nan];
    return;
end
delta = std(y)
avg = mean(y) it can be local std and mean with window equal to +-several values of "s"

[Bernard Grundlehner]

Too slow on large sets of data. I prefer peakdetect.m

[Frederick Lu]

Performance need improvements.

[Linda Best]

i found extreme.m is geat and also find peaks from surface: extreme2.m! But this was the first i used

[Rory Gamble]

I would really like to see a version of this that does functions of two wariables, eg extreme.m/extreme2.m

[Sunil Arya]

I found a very good m code to pick up the peak data.
Thankyou

[Mukesh V.]

this is the best that I found, relavent code

[Estella Liu]

This works great!! Thank you so much :)

[Vlad Vlad]

Excellent! Thank you

[Rishi Bansal]

works like a charm!!!

[Vinh Nguyen Quang]

 

[Kristin Golmon]

Doesn't work when there are two data matching values at the top of a peak. Other than that, perfect.

[shilin guo]

working perfect!!

[john Chandler]

it doesn't work.

Change line 9 in ma.m to
t=fpeak(A,B,30,[0,90,100,inf]);

and you will see that it picks peaks and valleys, both.

[antonio Acampora]

 

[David Jessop]

Poorly explained, poor performance for large data sets and requires too much parameter tweaking in order to find peaks.

[JQZhou]

There is a bug in this program. when the peak/valley contains more than 1 point with same value, this program will just ignore this peak/valley.

correction:
replace
tP=(sum(y(top:bottom)>=y(i))==1);
bP=(sum(y(top:bottom)<=y(i))==1);
with
tP=(y(ii)==max(y(top:bottom)));
bP=(y(ii)==min(y(top:bottom)));

[Alex]

Sorry if this is a dumb question, but where is this getPeak function that is called within this function?? I tried finding it on File Exchange and am not getting anything...

Thanks

[Alex]

Ah never mind... d'oh

[Caleb]

Great performance.

[Kami]

I would be grateful if you put some explanation about the inputs. What is the input "Range"? and why is it a four-element vector?

[Kami]

About the input parameter,"Range":
Range is a 4-element vector:
[Xmin, Xmax, Ymin, Ymax]
Xmin: the lower x boundary of region of interest(ROI).
Xmax: the upper x boundary of ROI
Ymin and Ymax are the same as Xmin and Xmax.
For example,
t=fpeak(A,B,30,[23,90,700,inf]);
means:
find the peaks whose x are between 23 and 90 and their values is between 700 and inf.