Sensing the data changes and plotting

I have a table "a" which has multiple rows and 2 columns; t for time and q for data. I attached here.

I want to plots the period only where the sudden change of value happens (from start to end of the changes) such as I showed in figure where the value jumped from nearly -4 to -43.65.

After that I'd like to generate a table of the data of the plot. There will be many periods where such kind of data changes occurs so I want to plot everytime with that data changes.

load = 'newdata3.csv';
data = readtable(load);
data = sortrows(data,'Var1','ascend');

timetable(data.Var1, data.Var2);
plot(data.Var1,data.Var2)

findchangepts(data.Var2,'MaxNumChanges',20)

Here, I tried and it is not showing at the correct changes points. There are 12 changes points and it is only showing 5 of them. I don't know how to fix this.

Answer 1

[Image Analyst]

Do you have the Image Processing Toolbox? If so, it's easy.

v = a.V;
mask = v > 200
props = regionprops(mask, 'PixelIdxList');
numRegions = numel(props);
coords = vertcat(props.PixelIdxList);
startingRows = coords(:, 2);

or something similar. It would have been easier if you had read the posting guidelines and remembered to attach your data.

To learn other fundamental concepts, invest 2 hours of your time here:

MATLAB Academy - Free 2 hour training

See my Image Processing Tutorial in my File Exchange:

Image Analyst's File Exchange

If you have any more questions, then attach your data and code to read it in with the paperclip icon after you read this:

TUTORIAL: How to ask a question (on Answers) and get a fast answer

[uzzi]

Thank you for your guideline. I watched that free 2 hours training. That was helpful. I took your advice and improved my question with the attachment. But your code is not working for my file. Can you take a look into it again with my attached csv file?

[Image Analyst]

Did you try findchangepts?

[uzzi]

Yes, I tried. It doesn't show any changes

[Image Analyst]

Your data looks weird:

a = readtable('newdata3.csv')

a = 126000×2 table
       Var1         Var2  
    ___________    _______

    12:14:54 AM    -3.3268
    12:14:54 AM    -4.1778
    12:14:54 AM    -4.6337
    12:14:54 AM     -4.729
    12:14:54 AM    -4.9851
    12:14:54 AM    -4.2736
    12:14:54 AM    -3.2882
    12:14:54 AM     -2.257
    12:14:54 AM    -1.5046
    12:14:54 AM    -1.6617
    12:14:54 AM    -1.4103
    12:14:54 AM    -2.4553
    12:14:54 AM    -3.3668
    12:14:54 AM     -4.193
    12:14:54 AM    -5.1044
    12:14:54 AM     -5.133

t = a.Var1;
v = a.Var2;
plot(t, v, 'b-');
grid on;

mask = v > 20;
props = regionprops(mask, 'PixelIdxList');
numRegions = numel(props)

numRegions = 310

startingRows = vertcat(props.PixelIdxList)

startingRows = 6462×1

54011 54013 54014 54015 54016 54017 54018 54019 54020 54021

[uzzi]

I think that's lack of table sorting. After I read the table in MATLAB, I sort the rows ascending with time. a = sortrows(a,'Var1','ascend'); Maybe that's why.

[Image Analyst]

OK, what's your threshold? It looks like you could use a global threshold everywhere, like 5 or something.

[uzzi]

The threshold is 20 and -20.

[uzzi]

Is it possible to use image processing commands for data processing?

[Image Analyst]

Yes, of course. Images are data. And many of the image processing functions work just as well on 1-D signals as well as 2-D and 3-D images.

[uzzi]

Are there anyway to generate graph automatically in the areas when big data changes happens?

[Image Analyst]

Yes. But I don't know what you want.

If you have any more questions, then attach your data and code to read it in with the paperclip icon after you read this:

TUTORIAL: How to ask a question (on Answers) and get a fast answer

[uzzi]

I think you didn't understand my question. That's the question I'm asking from the start. I can plot in conventional way. According to the plot, you can see that there are 6 places where big data changes happen.

What I want is I want the MATLAB to sense the data automatically and generate a graph where a big data changes is sensed.

In this case, I want the MATLAB to automatically generated 6 figures where big data changes occurs.

I hope this explanation is clear to you.