combine multiple text files into one text file

Hi

I have multiple text files with names like 1.txt, 2.txt, 3.txt...... Each file has a single row of data(same format and length), I am think about combining those files into one text file which has a matrix data.

For example

1.txt has a row of " 1 3 4 8...."

2.txt has a row of " 3 5 7 3...."

3.txt has a row of " 9 1 3 5...."

......

I want to combine them into a file with a matrix like below,

1 3 4 8....

3 5 7 3....

9 1 3 5....

...........

...........

Thanks a lot

[ranjith J]

source code need

Answer 1

[Fangjun Jiang]

In DOS, I used to use this all the time.

copy 1.txt+2.txt+3.txt MyBigFat.txt

In MATLAB, you can use

system('copy 1.txt+2.txt+3.txt MyBigFat.txt')

[Jason Ross]

Cool! I didn't know that would work ... but I'll file it in the "potentially useful" file.

[Matlab2010]

Doesnt work if writeStr or finalNameStr contains white spaces. Due to how DOS works

system(['copy ' writeStr ' ' finalNameStr]);

[BA]

I had the Problem that the final file contained an arrow as last sign which causes trouble. This was solved by using

system('copy/b ...');

Description for copy/b

[Robert]

This gives me a SUB character at the end of each appended file (https://en.wikipedia.org/wiki/Substitute_character). Do you know how to prevent this, or how to remove these? Regexping doesn't work as the files are too large.

[michael fried]

Use parameter "b" (binary) in order to avoid that character...:

In Windows Command Line:

copy /b 1.txt+2.txt+3.txt MyBigFat.txt 

Or in Matlab:

system('copy /b 1.txt+2.txt+3.txt MyBigFat.txt')

Answer 2

[Jason Ross]

Keep in mind you don't even need MATLAB to do this. You can do it with built in OS commands on Windows or UNIX

Windows:

type 1.txt > bigfile.txt
type 2.txt >> bigfile.txt

UNIX:

touch bigfile.txt
cat 1.txt >> bigfile.txt
cat 2.txt >> bigfile.txt

This will result in one big file that's in the format you want. Of course, there's no error checking, but the big upside is going to be that there is very little overhead so your processing may go considerably faster depending on the number of files you have to do.

Depending on the order of the file names, this may also be very easily scriptable. You will, of course need to use one of the import functions in MATLAB to read it in eventually, as well.

Answer 3

[Nirmal Gunaseelan]

I prefer to use TEXTREAD in such cases. You could pass in different file names in every call and use the same matrix with different row numbers as output variables.

[Cody]

I read textread help but still do not know how to use textread to combine them. Could you please write down the detailed steps? Thanks.

[Fangjun Jiang]

What Nirmal meant was to use textread() to read each file and then combine them in MATLAB. You know it's very easy to concatenate data in MATLAB, do you?

a=[1 3 4 8]

b=[3 5 7 3]

c=[a;b]

Answer 4

[Alex Z.]

This can be done in Easymorph (it's free) using Append transformation. It can work with as many as 10 millions rows in the file.