Combine many txt format files with a single column into one file

 files = {'A.txt', 'B.txt', 'C.txt', 'D.txt', 'E.txt'} ;
 buffer = cell( size( files )) ;
 for k = 1 : numel( files )
    buffer{k} = sscanf( fileread( files{k} ), '%f' ) ;
 end
 fId = fopen( 'Merged.txt', 'w' ) ;
 fprintf( fId, '%f\t%f\t%f\t%f\t%f\r\n', horzcat( buffer{:} ).' ) ;
 fclose( fId ) ;

You will have to tailor the formatSpec '%f\t%f\t%f\t%f\t%f\r\n' to your needs, i.e. define a more specific numeric format (e.g. %.3f instead of %f) and define the separator (here \t for tab, but you may need/want a simple comma instead).

5 个评论
显示 3更早的评论隐藏 3更早的评论

Cedric 2015-9-28

编辑：Cedric 2015-9-28

在 MATLAB Online 中打开

To add to what was already said, this is a string called format spec, which is passed to FPRINTF to define the format of what this function must output. It is a common way to define what and how to print something, or what and how to read something (S/FSCANF). It is usually used for printing variables content to the standard output (the command window), e.g.

 a = 8 ;
 b = 4.5 ;
 fprintf( '%d, %f\n', a, b ) ;

Output:

8, 4.500000

Here you see that there are two format operators in the format spec, %d (print integer) and %f (print float) that will be applied to the two variables that are passed to FPRINTF after the format spec in the same order ( %d will be applied to the content of variable a, and %f to the content of variable b). The last \n codes for new line.

This is pretty standard. Now the less standard part is illustrated below:

 a = [1,2,3,4] ;
 fprintf( '%d %d\n', a ) ;

Output:

 1 2
 3 4

Here you see that the format spec (output 2 integers and a new line) was repeated as many times as needed for "eating" ;-) all elements of array a. This allows to output many lines with a single call to FPRINTF and this is what I implemented in my solution. Note that the array is read column-wise and this is why we transpose the output of HORZCAT (n x 5 -> 5 x n).

sally_wu 2015-10-3