Rehsaping Data Table to new format

Hi, I have a data table that exists like:

[date] [field] [data]

1/1/2017 A 1

1/2/2017 A 2

1/3/2017 A 3

1/1/2017 B 3

1/2/2017 B 2

1/3/2017 B 1

1/1/2017 C 4

1/2/2017 C 5

1/3/2017 C 6

and i want to reshape the data to be like this:

[Date] [A] [B] [C]

1/1/2017 1 3 4

1/2/2017 2 2 5

1/3/2017 3 1 6

so that each unique value in fields becomes a new field name in the new table.

I saw this conversion done in a tutorial the other day, but cant recall how it was done at all, thanks for any input in advance!

Matt

Answer 1

[Peter Perkins]

I think you are looking for unstack:

>> t = cell2table( ...
    {'1/1/2017', 'A', 1; ...
     '1/2/2017', 'A', 2; ...
     '1/3/2017', 'A', 3; ...
     '1/1/2017', 'B', 3; ...
     '1/2/2017', 'B', 2; ...
     '1/3/2017', 'B', 1; ...
     '1/1/2017', 'C', 4; ...
     '1/2/2017', 'C', 5; ...
     '1/3/2017', 'C', 6},'VariableNames',{'date' 'field' 'data'});
>> t.date = datetime(t.date,'InputFormat','MM/dd/yyy');
>> t.field = categorical(t.field)
t =
  9×3 table
       date        field    data
    ___________    _____    ____
    01-Jan-2017      A       1  
    02-Jan-2017      A       2  
    03-Jan-2017      A       3  
    01-Jan-2017      B       3  
    02-Jan-2017      B       2  
    03-Jan-2017      B       1  
    01-Jan-2017      C       4  
    02-Jan-2017      C       5  
    03-Jan-2017      C       6  
>> t2 = unstack(t,'data','field')
t2 =
  3×4 table
       date        A    B    C
    ___________    _    _    _
    01-Jan-2017    1    3    4
    02-Jan-2017    2    2    5
    03-Jan-2017    3    1    6

Answer 2

[Steven Lord]

If your data is stored in a table or timetable, use unstack. The first example on that documentation page is pretty close to the exact scenario you gave, so it should be easy to adapt for your needs.

Answer 3

[Matt]

Both answer the question and accomplish what i was looking to do, thanks very much!