preprocessing data for PCA

Hi, I have a table contains categorial data as the first column and numerical data as the rest of columns. I need to fill the missing values and standardize the numerical data. Here is what I did:

PrepData = fillmissing(data,'spline','DataVariables',@isnumeric);
func = @zscore;
PrepData = varfun(func,PrepData{:,2:end});

but I receive this error:

Undefined function 'varfun' for input arguments of type 'function_handle'.

How can I fix this?

Thank you

Answer 1

[Shantanu Dixit]

Hi Mahmoud,

If I understood the query correctly, you're trying to preprocess a table where the first column is categorical and the remaining columns are numerical. The error occurs because 'varfun' expects a table as input: https://www.mathworks.com/help/matlab/ref/table.varfun.html#btyj4vl-1-A , but 'PrepData{:,2:end}' extracts a matrix (numeric array).

To fill the missing values in the numerical columns (using 'spline' interpolation) and correspondingly standardize the numerical data (using 'zscore'), you can refer to a below example script:

% Sample data  
categories = {'Type1'; 'Type2'; 'Type1'; 'Type2'; 'Type1'};  
values1 = [1.2; NaN; 3.4; 4.1; 5.6];  
values2 = [10.5; 20.1; NaN; 40.3; 50.7];  
data = table(categories, values1, values2, ...  
            'VariableNames', {'Category', 'Feature1', 'Feature2'});  

disp(data)

    Category     Feature1    Feature2
    _________    ________    ________

    {'Type1'}      1.2         10.5  
    {'Type2'}      NaN         20.1  
    {'Type1'}      3.4          NaN  
    {'Type2'}      4.1         40.3  
    {'Type1'}      5.6         50.7  

% fill missing values  
PrepData = fillmissing(data, 'spline', 'DataVariables', @isnumeric);  

% Standardize numerical columns (keeping categorical)  
numVars = PrepData.Properties.VariableNames(2:end); % Get numerical column names  

% varfun expects table /timetable as second input argument
PrepData(:,numVars) = varfun(@zscore, PrepData(:,numVars));  

disp('Preprocessed data:');  

Preprocessed data:

disp(PrepData);  

    Category     Feature1    Feature2 
    _________    ________    _________

    {'Type1'}     -1.3476      -1.2467
    {'Type2'}    -0.42879     -0.64327
    {'Type1'}           0    -0.016763
    {'Type2'}     0.42879      0.62651
    {'Type1'}      1.3476       1.2803

Hope this helps!