How to Iterate over the columns of a Matlab table?

How can I iterate over the columns of a Matlab table? The Python equivalent and assuming data was a pandas data frame would be:

    variable_names=data.keys() #get variable names
    for i in variable_names:   # iterate over and print all data using variable names
        print data[i]

But the corresponding for a Matlab table does not work:

    f='Vilar_data.xls';
    data=readtable(f);
    variables=data.Properties.VariableNames; %get variable names
    for i=variables,
        data(1:end,{i})                      %attemt to iterate over the data by column headings
    end

[Oleg Komarov]

You need to remove the brackets from {i}, because i will already be a cell containing a string.

Answer 1

[Geoff Hayes]

Ciaran - access data in a table illustrates a couple of different ways in which you can access the column data within the table. You can access the column either by its variable index or by its variable name. For example, if you wish to access each column by the variable index, you could something like the following

 for k=1:length(variables)
    colData = data.(k);
 end

In the above, we iterate over each variable and access the column data directly through the variable index.

[Oleg Komarov]

The eval is pretty bad, please remove it. You can also index a table as you would do with a structure, with string fieldnames.

[Maryam Abbaszadeh]

x = [];

x = ['A':'Z'];

for i='A':'Z'

for j='A':'Z'

x = [cellstr(x),[i,j]];

end

end

for i=1:100

app.UITable.ColumnName ={x(i)};

end

i want to change ColumnNames with x characters but it doesn't work with above code

how can I solve this problem?

[Image Analyst]

No idea what you want. Please read this link then post your question in a new question along with your desired output.

Answer 2

[Image Analyst]

I would not iterate over variable names, unless they might possibly be created by some weird program that changes what column the data were in. I'd simply just use column number - it's so so much simpler. Run this demo:

% Create sample data (from the help documentation).
LastName = {'Smith';'Johnson';'Williams';'Jones';'Brown'};
Age = [38;43;38;40;49];
Height = [71;69;64;67;64];
Weight = [176;163;131;133;119];
BloodPressure = [124 93; 109 77; 125 83; 117 75; 122 80];
% Create a table, T, as a container for the workspace variables.
T = table(Age,Height,Weight,BloodPressure,'RowNames',LastName)
% Let's see what methods T has
methods(T)

% Now let's iterate over all columns using column number.
for col = 1 : width(T)
  fprintf('\nHere is column #%d, which is called %s\n', col, T.Properties.VariableNames{col})
  thisColumn = T(:, col) % Extract this one column into its own variable.
  % Now do whatever you want with "thisColumn" variable.
end

[Ciaran]

Hi Image Analyst, yes, precisely. I'm trying to write a program that will take any table so I think iterating over the column headings is best. Thank you though, your answer is still very informative

[Image Analyst]

I don't think so. Plus it's way more complicated. Why do you think it's better? My code will work with any table just using a simple index like you're already used to with other variable types. I'm not sure why you imply my code won't "take any table." It certainly will, and in a lot more straightforward manner.

Answer 3

[Walter Roberson]

for i=variables
      data(1:end,i{1}) 
end