How to fill in missing data?

Hello everybody,

I have a dataset(txt file) which contains some missing values (represented with 0). I would like to replace all this 0 places with numbers. How can I do?

Answer 1

[Ayush]

Hi,

One way to replace missing values currently represented as 0 is to use interpolation. Start by identifying the indices of the zeros in your dataset. Then, apply the “interp1” function to perform the interpolation. Refer to the pseudo code below for a better understanding:

% Step 1: Load the data
data = readmatrix('your_dataset.txt');

% Step 2: Interpolate to replace zeros
for col = 1:size(data, 2)
    x = 1:size(data, 1); % Indices of the data
    y = data(:, col);    % Data values

    % Find indices of non-zero and zero elements
    nonZeroIndices = y ~= 0;
    zeroIndices = y == 0;

    % Perform interpolation only if there are non-zero elements
    if any(nonZeroIndices)
        % Interpolate only non-zero elements
        yInterpolated = interp1(x(nonZeroIndices), y(nonZeroIndices), x(zeroIndices), 'linear', 'extrap');
        
        % Replace zeros with interpolated values
        y(zeroIndices) = yInterpolated;
    end

    % Update the column in the dataset
    data(:, col) = y;
end

% Step 3: Save the modified data back to a file (optional)
writematrix(data, 'modified_dataset.txt');

For more information on using the “interp1” function, please refer to the documentation below:

https://www.mathworks.com/help/matlab/ref/double.interp1.html

Answer 2

[Star Strider]

If you have R2016b, use the fillmissing function (introduced in R2016b) —

T1 = array2table(randi([0 9],10,5))                                             % Original

T1 = 10x5 table
    Var1    Var2    Var3    Var4    Var5
    ____    ____    ____    ____    ____

     2       8       7       7       1  
     8       7       5       6       7  
     8       4       4       5       3  
     4       3       3       1       3  
     4       6       5       5       7  
     0       4       7       0       4  
     8       2       2       7       2  
     9       7       3       0       9  
     5       0       5       8       3  
     2       2       6       3       4  

loc = table2array(T1) == 0                                                      % Logical Matrix Of ‘0’ Locations

loc = 10x5 logical array
   0   0   0   0   0
   0   0   0   0   0
   0   0   0   0   0
   0   0   0   0   0
   0   0   0   0   0
   1   0   0   1   0
   0   0   0   0   0
   0   0   0   1   0
   0   1   0   0   0
   0   0   0   0   0

T1 = fillmissing(T1,  'linear', MissingLocations=loc, EndValues='nearest')      % 'linear' Interpolation Of Missing Values With 'nearest' For End Values

T1 = 10x5 table
    Var1    Var2    Var3    Var4    Var5
    ____    ____    ____    ____    ____

     2        8      7        7      1  
     8        7      5        6      7  
     8        4      4        5      3  
     4        3      3        1      3  
     4        6      5        5      7  
     6        4      7        6      4  
     8        2      2        7      2  
     9        7      3      7.5      9  
     5      4.5      5        8      3  
     2        2      6        3      4  

There are several methods to fill (interpolate) the missing values. See the documentation for those and other options.

.