Loop to merge every two columns

Renyulb28 has asked for the wisdom of the Perl Monks concerning the following question:

Hi monks, the dilemma that we are having right now is with figuring out how to merge every two columns within the data file. The file is as follows; first column is the ID, and every column after that is a marker, except two markers make a whole, so that is what we want. These are space delimited. The colleague who knows python only suggested a long difficult method of writing out each combination (ie 1+2, 3+4, etc.) and with R it is even more problematic.

3851 A A G G T T ...
3854 A A G G T T ...
[download]

The output for the above example would look like this:

3851 AA GG TT ...
3854 AA GG TT ...
[download]

Is there a simple Perl script available for this task?

Comment on Loop to merge every two columns Select or Download Code

Replies are listed 'Best First'.
Re: Loop to merge every two columns by wind (Priest) on May 11, 2011 at 20:30 UTC
`use strict; while (<DATA>) { s/( \w) /$1/g; print; } __DATA__ 3851 A A G G T T 3854 A A G G T T` [download] Or as a one liner `perl -pi.bak -e 's/( \w) /$1/g' file.dat` [download] And if you want to use "newer" regex features, the following is shorter: `s/ \w\K //g;` [download]	[reply] [d/l] [select]
Re^2: Loop to merge every two columns by JavaFan (Canon) on May 12, 2011 at 09:33 UTC
Note that those "newer" regex features are present in all versions of Perl that aren't "end-of-life". The last major version of Perl that didn't have those was 5.8, which dates from July 2002 - less than a year after the release of IE 6.	[reply]
Re^3: Loop to merge every two columns by wind (Priest) on May 20, 2011 at 01:45 UTC
Yes, I provide some programming work for a couple non-profits that run perl 5.8 and perl 5.10 respectively. As a consequence, I don't always get to utilize the latest and greatest. And even though the 5.10 EOL was announced, I find that a lot of posters aren't always on a current version of perl. Therefore sometimes it's easier to just assume they don't have access to certain features. I definitely appreciate the \K marker though, and use it plenty now to clean up regex's when I can.	[reply]
Re^4: Loop to merge every two columns by JavaFan (Canon) on May 20, 2011 at 11:14 UTC
Re^5: Loop to merge every two columns by Tanktalus (Canon) on May 20, 2011 at 21:00 UTC
Some notes below your chosen depth have not been shown here
Re^5: Loop to merge every two columns by stvn (Monsignor) on May 21, 2011 at 01:31 UTC
Re^5: Loop to merge every two columns by wind (Priest) on May 20, 2011 at 17:47 UTC
Re: Loop to merge every two columns by LanX (Saint) on May 11, 2011 at 20:33 UTC
`DB<104> $a="3851 A A G G T T" DB<105> $a=~s/([ACGT]) ([ACGT])/$1$2/g DB<106> p $a 3851 AA GG TT` [download] Cheers Rolf PS: > The colleague who knows python only suggested a long difficult method of writing out each combination (ie 1+2, 3+4, etc.)... This pretty much reflects the impression I have when reading in python boards... :)	[reply] [d/l]
Re: Loop to merge every two columns by johngg (Canon) on May 11, 2011 at 22:57 UTC
A solution using regular expressions and substitution, as already shown, is probably best but just to show another way - `knoppix@Microknoppix:~$ perl -E ' > $_ = q{3851 A A G G T T}; > say; > @e = split m{( )}; > $_ = join q{}, @e[ grep { ( $_ + 1 ) % 4 } 0 .. $#e ]; > say;' 3851 A A G G T T 3851 AA GG TT knoppix@Microknoppix:~$` [download] I hope this is of interest. Cheers, JohnGG	[reply] [d/l]