How to find column where 0 turns to 1 for each row

Hello!

I would like to be able to find the column (or column title) for each row where the 0 turns to 1.

I have attached example rows (egcolumnsep.mat), so for example for row 5, I would like it to return column 20 (or Bins_11). I would like an output for all the rows though. Some rows might have only 0's (this can be NaN). This basically gives me the reaction time for each row.

Thank you for your help! =)

Answer 1

[Bruno Luong]

load('egcolumnsep.mat');

c1 = find(strcmp(DATAb1.Properties.VariableNames,'Bins_1')); % == 10
% Extract relevant data
A = table2array(DATAb1(:,c1:end));
if ~all(ismember(A,[0 1]),'all')
    error('Data must be 0 or 1')
end

[cols,rows] = find(diff(A.')==1);
cols = cols+c1;
Idx01 = table(rows,cols)

Idx01 = 37×2 table
    rows    cols
    ____    ____

     1      468 
     1      499 
     1      855 
     1      890 
     2      196 
     3      452 
     3      487 
     3      531 
     3      828 
     3      860 
     4      306 
     4      338 
     4      370 
     4      404 
     4      434 
     4      468 

[Tanika Bawa]

Thank you! Do you know how I could get it so that I only get ther first transition from 0 to 1?

[Bruno Luong]

Why didn't tell us that you are interested only the first transition in the question?

You can call this statement after the above scriot

groupsummary(Idx01, 'rows','min')

or start from scratch

load('egcolumnsep.mat');
c1 = find(strcmp(DATAb1.Properties.VariableNames,'Bins_1')); % == 10
% Extract relevant data
A = table2array(DATAb1(:,c1:end));
if ~all(ismember(A,[0 1]),'all')
    error('Data must be 0 or 1')
end

[a,cols] = max(A,[],2);
cols = (cols+c1-1) .* (a==1) % returns 0 if no transition 0->1 is found

cols = 5×1

468 196 452 306 20

Answer 2

[Jonas]

yourMat=[0 0 0 0 0;
    0 1 1 0 0;
    0 0 0 1 0];
where=arrayfun(@(idx)find(diff(yourMat(idx,:))==1,1,'first')+1,1:size(yourMat,1),'UniformOutput',false)'

where = 3×1 cell array
    {1×0 double}
    {[       2]}
    {[       4]}

an empty cell shows that there was not such a change

you could set it to NaN and convert to an normal array by

where{cellfun('isempty',where)}=NaN

where = 3×1 cell array
    {[NaN]}
    {[  2]}
    {[  4]}

where=cell2mat(where)

where = 3×1

NaN 2 4

[Tanika Bawa]

Hey it gives me this error when I try to use it

"Check for missing argument or incorrect argument data type in call to function 'diff'."

I think because yours is a double and mine is a table

[Jonas]

you can convert your table to a matrix before, e.g. yourTable{:,:} or table2array()

[Tanika Bawa]

I did try that and it doesn't really let me do it unfortunately. It gives me this error:

"Unable to concatenate the specified table variables.

Caused by:

Error using categorical/cat (line 56)

Unable to concatenate a double array and a categorical array."

I do need to keep all the columns (because in the end I just want to add a column at the end with the results from this)