how to label the variance decomposition table

Currently I have this code where I have 7 predictors Age, Weight, ...MaxPulse where Oxygen is the response variable

which gives the following output

However I want to modify the code such that in place of Var1, Var2, Var3, ....Var7 I have the predictors Intercept, Age, Weight, ....,MaxPulse.

Answer 1

[Paul]

Check the doc page collintest and review how to use the VarNames input. Or, instead of pulling all the variables out from data, and then stuffing those into an array, why not take advantage of the fact that collintest an accept a table as an input, and the DataVariables argument indicates which table variables are of interest.

collintest(data,DataVariables=["x0" "Age" etc])

Read the linked doc page for more information. Feel free to post a comment to this answer with the code if still having a problem.

[N/A]

Hello Paul, when I tried doing this

I get the output

[Paul]

The VarNames argument is used if the input data is an array. The DataVariables argument is used if the input data is a table.

For the former

collintest(Predictors,VarNames=["Intercept", "Age", etc ])

For the latter

collintest(data,DataVariables=["Intercept", "Age", etc])

Again, read the doc page collintest to better understand those two usages. It's possible that the output arguments might be different as well.

Answer 2

[Umar]

Hi Nafisa,

You asked, However I want to modify the code such that in place of Var1, Var2, Var3, ....Var7 I have the predictors Intercept, Age, Weight, ....,MaxPulse.

To answer your question, I need to update the variable names in the regression model. Below is a detailed example along with an explanation of how to accomplish this task:

Let us generate sample data with 100 samples and 7 predictors.

>> % Sample data for demonstration data = randn(100, 7); % Assuming 100 samples and 7 predictors response = randn(100, 1); % Response variable (Oxygen)

Then, create a table predictorTable with predictor variables named 'Intercept', 'Age', 'Weight', 'Var4', 'Var5', 'Var6', and 'MaxPulse'.

>> % Create a table with predictor variables predictorTable = array2table(data, 'VariableNames', {'Intercept', 'Age', 'Weight', 'Var4', 'Var5', 'Var6', 'MaxPulse'});

Then, fit a linear regression model lm using the fitlm function with the predictor table and the response variable.

>> % Fit a linear regression model lm = fitlm(predictorTable, response);

Finally, display the regression results using disp(lm).

>> % Display the regression results disp(lm);

Hope this will help resolve your problem.

[Paul]

How does fitlm relate at all to this question?

[Umar]

Hi Paul,

It effectively uses specific predictor names in linear regression model. Hence, the reason. But it sounds like you have provided your input as well which is well appreciated. Thanks for your contribution towards resolving this problem.