How to get the point index of the position of the red line in the waveform graph

[Chihl]

How to get the point index of the position of the red line in the waveform graph

[daveTW]

Use "Map Coordinates to XY" - Invoke node

[Chihl]

What I mean is to find this feature point through an algorithm, such as derivative , etc. Can you give me an example based on a waveform diagram?

[daveTW · Accepted solution]

Like this?

[altenbach]

If you create a cursor that is locked to a plot, the cursor index property will give you the index of the array element at that position. (only works if the cursor is locked to a plot, of course!)

You asked for the "point Index", not the x or y value. Right? If not, please be very specific!

[daveTW]

It didn't come into my mind, that the TO wants to find that point manually, but you are right - probably this was the question.

Because of the very unspecific question I was assuming that he wants to analyze a monotonic falling signal with a negative peak and wants to find the index of the smallest value within the negative peak.

If the signal doesn't meet these assumptions, the algorithm will fail.

Edit: No, he was asking for an algorithm

[altenbach]

---

@daveTW wrote:

Edit: No, he was asking for an algorithm

---

The red angled line (not even an arrow!) is partially on the graph and partially on the diagram and this does not point at any "index". The end of the red line does not even point at any particular point of the blue curve, but points at some nowhere land below a dip and the next hill. The red line is just a decoration and has no "index". Each point of the blue line has an index.

The curve looks like a very low order polynomial soon after, so maybe they want to find out where the glitching ends and the smooth curve starts. That would be simple.

---

@daveTW wrote:

Edit: No, he was asking for an algorithm

---

Well, the sentence "What I mean is to find this feature point through an algorithm, such as derivative , etc." does not really narrow it down because there are many features and the red line does not point at anything in particular.

Let's wait for detailed clarifications before we waste even more time guessing. 😄

[altenbach]

---

@Chihl wrote:
What I mean is to find this feature point through an algorithm, such as derivative , etc. 

---

If you do a least absolute residual (insensitive to outliers!) third order polynomial fit of all data (or a least square fit of most of the data, i.e. the last 90%), you can easy tell from the difference where "features" are, whatever that means.

[Chihl]

Thank you for your enthusiastic replies, I have found the answer I was looking for in so many replies. I've accepted the first answer that I think fits my idea, thanks again! !

[altenbach]

---

@Chihl wrote:
 I've accepted the first answer that I think fits my idea, thanks again! !

---

You can also mark more than one post as answer if applicable. Nothing wrong with that..