Truncating Y values during integration to integrate only the "peaks" or delete CERTAIN elements from a matrix

I have a load of data from a force transducer output. The data is in the form of time vs volts.

      TEXT=fopen('Test 5.txt','rt');
      texta=textscan(TEXT,'%f%f','CommentStyle','"');
      A=texta{1};
      B=texta{2};
      C=[A,B];
      D=C;
      plot(D)

The first part of my script removes the -ve values, as this relates to the compression wave recoiling in the force transducer and doesn't actually relate to compressive forces at all.

          C(C<0)=0;
          plot(C)

So far so good.

I am measuring the peak impulse above a certain range — impulse being the integral of force*time (or volts*time in this instance as the transducer outputs a voltage and I haven't included the conversion yet).

To measure the peak impulse I only need the values above a certain range, say 1.

I have tried to simply make a conditional statement to remove all data less than 1:

          E=C;
          E(E<1)=0;
          plot(E)

However, this is not correct as what I am after is more like this (edited on MS paint):

Does anyone have any ideas on how I can, either:

• Limit the Y values used to numerically integrate the data i.e ymin=1
• Remove the data in such a fashion as to resemble the final image above?

Any help would be appreciated!!! Mark.

EDIT:

Please note that I cannot simply state that E(E<1)=1 as this would artificially inflate the integral. I need all the data below that point to basically disappear.

Answer 1

[Azzi Abdelmalek]

findpeaks(y,'minpeakheight',1)

[mark]

Hey Azzi, you are a actual God!

It looks to be working, but I wont know for sure until I compare impulse values with that we expect from the rig drilling progress. Thank you!

I'll let you know.

[mark]

Hey Azzi,

It is able to take out values of y above 1 but it fails to take the X values too.

Say a force parabola (peak) is bounded within 0.1s, it will only plot the parabola force (i.e the line) vs a row value. If there are 100 points defining the curve it'll plot the Y values vs X values going from 0->100. It should take the corresponding time interval that the peak lies within.

As you can see in the above output, there are over 800 points that define data points above 1 but it hasn't indicated the total time over which this should occur.

• How do you get matlab to output the corresponding X values?
• Is there really no way to just tell matlab to integrate the area under the curve starting from this value?

cheers, Mark.

[Azzi Abdelmalek]

[x_peak,x_location]=findpeaks(y,'minpeakheight',1)

to integrate the area under the curve, use trapz function