Calculate multiple means in table by column indices

Hello community,

I want to calculate means of several parameters in my table according to row indices. Here is what my table now looks like:

    Subject                Phase                  Time           Par1             Par2      Par3              ... 
    __________________    ___________________    ____________    _____________    ______    _____________     _____________

    {'Name1'         }    {'Phase1'         }    {'00:17:05'}       1614.7        1.1787       136.33         ...
    {'Name2'         }    {'Phase1'         }    {'00:17:05'}        496.2         1.091       35.379         ...
    {'Name1'         }    {'Phase2'         }    {'00:15:05'}       1535.4        1.2158       127.32         ...
    {'Name2'         }    {'Phase2'         }    {'00:15:05'}       1084.9        1.1678       82.798         ...
   ...  % 98x9 table
  

The output should be a table:

      Phase                  Time           Par1             Par2      Par3              ... 
   ___________________    ____________    _____________    ______    _____________     _____________

   {'Phase1'         }    {'00:17:05'}    % mean values of all subjects in Phase1 ...
   {'Phase2'         }    {'00:15:05'}    % mean values of all subjects in Phase2 ... 
  
   ...  % 98x9 table
       

I calculated the mean with varfun of a single parameter (see underneath). However, I think there has to be a simple solution calculating the mean of multiple parameters without using for-loops and merging the output...

NewTable=varfun(@nanmean,Table,'InputVariables','Par1','GroupingVariables','Phase');

Answer 1

[Guillaume]

newtable = groupsummary(yourtable, {'Phase', 'Time'}, 'mean', 4:width(yourtable)) %mean of all but the first 3 variables, grouped by Phase and Time

Hopefully your table is not called Table but something more descriptive.

[Robin Schäfer]

Thank you for your very fast response! Works fine! I've renamed everything for this post and names are more precise in my functions =)