"Gregory" wrote in message <email@example.com>... > I have a table of data stored as a cell array that I need to sort into an order based on part of one the rows. > > Here is the table: (called "Table") > > 'MAR - Groups Total' 'CE1' 'CE2' 'CE3' 'CE5' 'CE6' 'CE7' 'CE8' 'Total' > 'Forecast' [27.6900] [62.1000] [88.0900] [62.4000] [73.5400] [ 94.2500] [80.2600] [488.3300] > 'Actual' [29.0600] [69.8500] [92.9400] [61.2700] [82.9100] [112.8000] [82.6100] [531.4400] > 'Difference' [-1.3700] [-7.7500] [-4.8500] [ 1.1300] [-9.3700] [-18.5500] [-2.3500] [-43.1100] > > What I want to do is ignore the first and last column, which contain the row titles and the totals, and sort by the absolute value of row 4. Then re order the rest of the rows based on the results of the sorting. > > for example. The lowest absolute difference is for group 'CE5' in column 5, so this should move to the start, column 2. The next smallest is 'CE1' which should be in column 3, then 'CE8' which should be in column 4 etc. > > I can do: > sort(abs(cell2mat(Table(4,2:end-1)))) > which sort into the correct order, but then how do I resort the rest of the data in the table based on this. > > Thanks for you help!
Sort can also output an index of the element locations before sorting. You can use this index to arrange the other rows in a similar manner:
[B,IX] = sort(A,...) also returns an array of indices IX, where size(IX) == size(A). If A is a vector, B = A(IX). If A is an m-by-n matrix, then each column of IX is a permutation vector of the corresponding column of A, such that