DIfferentiating repetitions in a repeated sequence

Hi,

I am working with eye tracking data. I have a variable that represents trials, that is made of a repeated sequence. It looks like this: [1,1,1,1,…,2,2,2,2,…,3,3,3,3,3,…,4,4,4,…] etc. I put dots here to show that there are a lot of these numbers. The sequence repeates itself each time a new bloc started in the experiment, and I would like to be able to know which bloc a trial belongs to.

How could I, for example, add 1000 to the first repetition of the sequence, then 2000 to the second one, so that I would know, if I see [3034] that I have the 34th trial from the 3rd bloc ?

Thank you for your help!

Answer 1

[Ameer Hamza]

Try this

x = [1 1 1 1 2 2 2 3 3 3 3 3 1 1 2 3 3 1 2 3];
idx = [1 find(diff(x)<0) + 1];
y = zeros(size(x));
y(idx) = 1;
y = cumsum(y);

res = splitapply(@(xo, yo) {splitapply(@(xi) {1000*yo(1) + xi(1)*ones(size(xi))}, xo, xo)}, x, y, y);
res = [res{:}];
res = [res{:}];

Result

>> res
res =
  Columns 1 through 6
        1001        1001        1001        1001        1002        1002
  Columns 7 through 12
        1002        1003        1003        1003        1003        1003
  Columns 13 through 18
        2001        2001        2002        2003        2003        3001
  Columns 19 through 20
        3002        3003

[Timothee Maniquet]

Hi, thank you very much for your answer.

What you suggest works perfectly, however, given an input

x = [1 1 1 1 2 2 2 3 3 3 3 3]

What I would like to obtain would be:

y=[ 1001 1001 1001 1001 1002 1002 1002 1003 1003 1003 ]

Such that the next time a '1' arrives in x, the ouput in y shows 2001.

Does that make sense? Thank you for the help.

[Ameer Hamza]

Try this

x = [1 1 1 1 2 2 2 3 3 3 3 3 1 1 2 3 3 1 2 3];
idx = [1 find(diff(x)==-2) + 1];
y = zeros(size(x));
y(idx) = 1;
y = cumsum(y);

res = splitapply(@(xo, yo) {splitapply(@(xi) {1000*yo(1) + xi(1)*ones(size(xi))}, xo, xo)}, x, y, y);
res = [res{:}];
res = [res{:}];

[Timothee Maniquet]

Thanks for your help. Unfortunately, this does not work on my data; in fact, when I run these lines using your example x = [1 1 1 1 2 2 2 3 3 3 3 3 1 1 2 3 3 1 2 3], I obtain:

y = 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 3 3 3

Do you obtain the same results, or I am doing something wrong?

[Ameer Hamza]

Check the value in variable ‘res’.

[Timothee Maniquet]

Ha of course! I hadn't checked 'res', my bad. So I did, but: 'res' contains the same numbers of values as 'x', however, it starts by '1001' and ends with '1054'. It seems like 'x' was order by value, pushing all the separate repetition of the [1,2,3,…] sequence together, and then applying the '+1000' operation, is that possible ?

Thank you for your help!

[Ameer Hamza]

When i run these lines of code

x = [1 1 1 1 2 2 2 3 3 3 3 3 1 1 2 3 3 1 2 3];
idx = [1 find(diff(x)==-2) + 1];
y = zeros(size(x));
y(idx) = 1;
y = cumsum(y);
res = splitapply(@(xo, yo) {splitapply(@(xi) {1000*yo(1) + xi(1)*ones(size(xi))}, xo, xo)}, x, y, y);
res = [res{:}];
res = [res{:}];

I get following output

>> res
res =
  Columns 1 through 6
        1001        1001        1001        1001        1002        1002
  Columns 7 through 12
        1002        1003        1003        1003        1003        1003
  Columns 13 through 18
        2001        2001        2002        2003        2003        3001
  Columns 19 through 20
        3002        3003

Can you show how this should be changed?

[Timothee Maniquet]

This is strange. When I run the lines of code you sent, using x = [1 1 1 1 2 2 2 3 3 3 3 3 1 1 2 3 3 1 2 3]; I have the same output. However, when I run it on my data, even after controlling that it contains the repeated sequence as I expect them, I obtain the same 'res' which doesn't go further than 1054.

So my input 'x' here is a 1X143864 double, that contains something similar to this sequence:

x = 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 … 54

Except that each number is repeated many times (a 1 here would be 500 1s in the real data), and the sequence repeats itself many times.

So when I run :

idx = [1 find(diff(x)==-2) + 1];

y = zeros(size(x));

y(idx) = 1;

y = cumsum(y);

res = splitapply(@(xo, yo) {splitapply(@(xi) {1000*yo(1) + xi(1)*ones(size(xi))}, xo, xo)}, x, y, y);

res = [res{:}];

res = [res{:}];

On my 'x', I obtain a 'res' of the same size, but that ends with:

… 1054 1054 1054

It's as if all the values had been ordered numerically before applying the operation. I don't know what to do with this.

Thanks for the help!

[Ameer Hamza]

Can you give an example matrix 'x' which does not work as expected? Attach it in a .mat file.

[Timothee Maniquet]

Sure, here it is.

[Ameer Hamza]

Check the updated code in my answer.

[Timothee Maniquet]

It worked! Thank you so much for your help.

[Ameer Hamza]

I am glad to be of help!