Find The Relationship between Data Set

15 Ansichten (letzte 30 Tage)
Arygianni Valentino
Arygianni Valentino am 27 Feb. 2018
Bearbeitet: Pawel Jastrzebski am 27 Feb. 2018
Right now I have 10 data set, with total around 800 times measurement (so I would have matrix with dimension 800x10), I want to know the relationship between each column of my dataset.
Is there any matlab toolbox that I can use to handle this problem?
Thank you for your time before.
  2 Kommentare
jonas
jonas am 27 Feb. 2018
Statistics
see e.g. corr() function
Arygianni Valentino
Arygianni Valentino am 27 Feb. 2018
that is what I looking for. Thank you very much for your respon.

Melden Sie sich an, um zu kommentieren.

Antworten (1)

Pawel Jastrzebski
Pawel Jastrzebski am 27 Feb. 2018
Bearbeitet: Pawel Jastrzebski am 27 Feb. 2018
Look into the following fucntions:
It should help you quickly assess visually whether the values in your matrix are correlated with each other. This way you can determine that:
i.e. column 1 and 5 are correlated but 1 and 4 aren't. Then carry out more elaborate data analysis on the identified sets of data.
Corrcoeff will help you quantify your findings (it will give you a level of correlation)
Consider the following example:
% memory preallocation
data = zeros(30,4);
% data generation
noise = 1-2*rand(30,1);
data(:,1) = 1:30;
data(:,1) = data(:,1) + noise;
data(:,2) = sin(1:30);
data(:,3) = 1 - cos(1:30);
noise = 1-2*rand(30,1);
data(:,4) = sqrt(data(:,1));
% matrix plotting
plotmatrix(data)
cVal = corrcoef(data)
The output is:
And
cVal =
1.0000 -0.1413 0.0376 0.9774
-0.1413 1.0000 -0.0167 -0.1641
0.0376 -0.0167 1.0000 0.0397
0.9774 -0.1641 0.0397 1.0000
Conclusion:
Seems like data from column 1 and 4 are correlated.

Kategorien

Mehr zu Probability Density Functions finden Sie in Help Center und File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by