Pearson correlation graph between variables

Hi

There are 6 variables (10,000 of each) which I am importing from excel and then using these, I am calculating a new variable let us say, Y (again, 10,000 iterations). How do I plot Pearson correlelation coeffcient between Y and each of 6 variables as tornado plots?

Thanks in advance

Answer 1

[Shubham]

Hi Sunil,

To create a tornado plot of Pearson Correlation Coefficients between a dependent variable ( Y ) and six independent variables, follow these steps:

1. Use the "corr" function to calculate correlation between ( Y ) and each independent variable. Below is the MATLAB code snippet to achieve this:

% List of variable names
variables = {'Var1', 'Var2', 'Var3', 'Var4', 'Var5', 'Var6'};
correlations = zeros(1, length(variables));

% Calculate correlations
for i = 1:length(variables)
    correlations(i) = corr(data.(variables{i}), Y);
end

2. Use the "barh" function to create a horizontal bar chart for the tornado plot.

barh(correlations, 'FaceColor', [0.5, 0.7, 1]);

Refer to the following MathWorks documentation link for more information on "corr":

https://mathworks.com/help/releases/R2024a/stats/corr.html

Hope this helps.

[Sunil]

Hi Shubham

Thanks for reply. Suppose out of 6, I want to plot the correlation of output "Y" with any 3-4, then how can I do that?

Thanks

[Sunil]

When I run the suggested program, I get the eroor:

Unable to resolve the name 'data.C'.

Error in MC_inputdist (line 156)

correlations(i) = corr(data.(variables{i}), HI);

what could be reason for this?

Thanks

[Shubham]

Hi Sunil,

This issue may be due to Field Names mismatch. Ensure that your "data" structure actually contains a field named "C".

You can check the fields of "data" using:

fieldnames(data)

Hope this helps.

[Sunil]

Hi Shubham

Thanks a lot! It worked.

Answer 2

[KSSV]

clc; clear ;
R = zeros(6,1) ;
for i = 1:6
    X = rand(100,1) ; 
    y = rand(100,1) ; 
    R(i) = regression(X',y') ; 
 
end
barh(R)

[Sunil]

I want to find Pearson correlation not regression relationships. Thanks