How to generate sequences from two vectors using rowfun

Hi,

I am trying to generate sequences for which starting values and last values+1 are contained in vectors, without using for-loop. Because there is a lot of data to process, doing for-loop is unfortunately not feasible. As the output, I want to have all the sequences to be concatenated into one vector. I think rowfun can help me here, but so far I have not been successful. If someone can give me a helping hand, I would be grateful.

This is an example;

x = [1,6,12]'; y = [5,10,13]';
the output I wish to obtain: z = [1,2,3,4,6,7,8,9,12];

I imagine it should be something like;

inputvecs = table(x,y);
z = rowfun(@seqgen,inputvecs)

How should the function seqgen look in this case? Should concatenation within the function?

Thank you very much for your help.

Etsuko

Answer 1

[Stephen23]

You could use arrayfun instead:

>> x = [1,6,12]; y = [5,10,13];
>> cell2mat(arrayfun(@(a,b)a:b-1, x,y,'UniformOutput',false))
ans =
    1    2    3    4    6    7    8    9   12

or

>> tmp = arrayfun(@(a,b)a:b-1,x,y,'UniformOutput',false);
>> horzcat(tmp{:})
ans =
    1    2    3    4    6    7    8    9   12

Note that the inputs must be horizontal vectors.

[Etsuko]

Thank you so much for your help, Stephen and Andrei!

Here I compared time for these three methods;

>> x = [1,6,12]; y = [5,10,13];

>> tic; cell2mat(arrayfun(@(a,b)a:b-1, x,y,'UniformOutput',false)); toc

Elapsed time is 0.008368 seconds.

>> tic; tmp = arrayfun(@(a,b)a:b-1,x,y,'UniformOutput',false);horzcat(tmp{:}); toc

Elapsed time is 0.000757 seconds.

>> tic; a = (min(x):max(y))';[ii,~] = find(bsxfun(@ge,a,x)&bsxfun(@lt,a,y));out = a(ii);toc

Elapsed time is 0.003983 seconds.

Thank you!

[Stephen23]

@Etsuko: note that to get meaningful timing values for such quick functions, it is recommended to repeat the process a few hundred or thousand times.

Answer 2

[Andrei Bobrov]

x = [1,6,12]; y = [5,10,13];
a = (min(x):max(y))';

[ii,~] = find(bsxfun(@ge,a,x)&bsxfun(@lt,a,y));
out = a(ii);

[Etsuko]

Thank you very much! It worked.