Automate data delimitation process to plot a range where the graph behavior is constant (resembles a straight line)

[mariajcuellar]

Hi, I am student and I am learning to use labview, but it has been difficult. So I need to cut off some txt docs which contain a range of data, but it has to be automatic. So I can't put the length and index in the flowchart program. I have to read 20 files and each one has another 6 documents inside. I figured out how to read these, but I am struggling with the Automation process to find the "useful part of the data" or to limit the noise in the graph. I am really confused So If someone could help me I will appreciate it so much. I attached my doc. The automation part must be replace the  enclosed part. Thanks!

[RavensFan]

The key part is figuring out what is useful and what is not.

You haven't told us anything about your data or what rules you have that determines what is useful.

[mariajcuellar]

Hi, You right, the process is basically that I have a bunch of data, when I plot it in some part of the graph the behavior is similar a straight horizontal line, the range is aprox between 1500 and 19500. Taking in account the last signal. So I have to cut it off in these point automatically but I don't know how. In the graph you can see the part that I have to preserve. Thanks for your help

[GerdW]

Hi maria,

---

@mariajcuellar wrote:

In the graph you can see the part that I have to preserve.

---

So you want the part of the signals when they are above 0.25?

Test when the signal rises above 0.25, then search for the index where they signals fall below 0.25.

Use the found indices for ArraySubset…

[RavensFan]

Or do you want the parts of the signals that are relatively flat?

If you do a level of filtering on the data to smooth it, then take the derivative (point by point), the flatness ends right about where the derivative dramatically increases (actually decreases since it is negative).