Import Text File Data With Blank Rows

I am having trouble importing a data set from a text file with blank rows. The data is in file "set1.txt" (attached) formatted as follows:

X data    Y data
-----------------
   1       5
   2       7.8
   3       2.1

X data2   Y data2
-----------------
   1       2
   2       2.4
   3       8

When I use

G=importdata('set1.txt');

I have the first set of data contained in

>>G.data
ans= 
   1       5
   2       7.8
   3       2.1

Everything below the blank (empty) row is ignored. What I'd ultimately like to do is import each chunk of data into either its own n x 2 matrix, or, into an appended matrix such that all the data in set1.txt is imported into a matrix with the following format:

     1       5       1   2
     2       7.8     2   2.4
     3       2.1     3   8

Any help would be much appreciated!

[per isakson]

See http://www.mathworks.com/matlabcentral/answers/127779#answer_135208, especially the comments

Answer 1

[Cedric]

Run the following

 content = fileread( 's1.txt' ) ;
 blocks  = regexp( content, '-[\n\r]+([^X]*)', 'tokens' ) ;
 blocks  = [blocks{:}] ;

and look at the content of blocks{1} and blocks{2}. Then see Per's answer in the thread that he links above. You can proceed the same way; I just wanted to provide you with a pattern which matches your setup, as it can be tricky to build if you are not familiar regular expressions.

[Nathanael]

Thanks guys. Between the two of you I got it working. I'm not familiar with these expressions, not a straight forward solution but it works! Thanks.

[Cedric]

The pattern defines:

• match the dash char: -
• followed by one or more char that is either a line break or a carriage return: |[\

]+|

• then take as many chars as possible which are not X: [^X]*
• and extract them as a token: ()

The regexp engine matches this pattern, extracts the token, and then goes on matching and extracting until the end of the content. While doing that, it stores tokens in a cell array. As a match can contain multiple tokens, the output cell array is a cell array of cell arrays. Each cell contains a cell array of all tokens related to a match. In your case, there is only one token per match, so each internal cell array contains only one cell. This is why I "flatten" blocks afterwards, so you end up having a cell array of tokens (instead of a cell array of cell arrays of tokens ;-))