Question

I have the following script for importing text files into matlab which include hourly data, where I am then trying to convert them into daily averages:

  clear all 
  pathName = ...
  TopFolder = pathName; 
  dirListing = dir(fullfile(TopFolder,'*.txt'));%Lists the folders in the directory specified by pathName.

  for i = 1:length(dirListing);
      SubFolder{i} = dirListing(i,1).name;%obtain the name of each folder in
              %the specified path.
  end

  %import data
    for i=1:length(SubFolder);
        rawData1{i} = importdata(fullfile(pathName,SubFolder{i}));
    end


  %convert into daily averages
    rawData2=cell2mat(rawData1);
        %create one matrix for entire data set
    altered=reshape(rawData2,24,(size(rawData2,2)*365));
        %convert into daily values
    altered=mean(altered)';
        %take the average for each day
    altered=reshape(altered,365,size(rawData2,2));
        %convert back into original format

My problem lies in trying to convert the data back into the same format as ‘rawData1’ which was a cell for each variable (where each variable is denoted by ‘SubFolder’. The reason for doing this is that all but one of the variables are vectors, where the remaining variable is a matrix (8760*11).

So, an example of this would be:

  clear all 

  cell_1 = rand(8760,1); 
  cell_2 = rand(8760,1);
  cell_3 = rand(8760,1);
  cell_4 = rand(8760,1);
  cell_5 = rand(8760,1);
  cell_6 = rand(8760,11);
  cell_7 = rand(8760,1);
  cell_8 = rand(8760,1);
  cell_9 = rand(8760,1);

  data = {cell_1,cell_2,cell_3,cell_4,cell_5,cell_6,cell_7,cell_8,cell_9};

Where I need to convert each cell in ‘data’ from hourly values into daily averages (i.e. 365 rows).

Any advice would be much appreciated.

Answer 1

I think this does what you want.

data = cellfun(@(x) reshape(mean(reshape(x,24,[]))',365,[]),data,'uniformoutput',false);

However that is kind of confusing so I will explain a little.

This part mean(reshape(x,24,[]))' inside of the cellfun will reshape each cell in data into a 24 by 365, compute the mean, then turn it back into a single column. This works fine when the original data only has 1 column … but for cell_6 with 11 columns it puts all the data end to end. So I added an addition reshape(...) wrapper around the mean(...) part to put it back into the original 11 columns … or more precises N columns that are 365 rows in length.

Note: This is going to give you errors if you ever have data sets dimensions are not 8760 by X.