MATLAB Answers


Read .txt and Write in matab

Asked by vaya putra on 11 Nov 2019 at 14:04
Latest activity Commented on by Rik
on 13 Nov 2019 at 17:02
i have 30000 thousand line of input file in text file, but i only need line 2618-3400.
imposible if i have to write manual to the end
how to write value C that consist of data from 2618-3400
%line 2618 to 3484
C =[A{2618,1} A{2619,1} A{2620,1} A{2621,1} A{2622,1} A{2623,1} A{2624,1}]
format long g
pf = sscanf(C,'%f')


Adam Danz
on 11 Nov 2019 at 15:43
"i have no idea with fileread(),"
It's really easy to read in the entire file. You literally just provide the filename. Once you've read in the entire file, then you just use indexing to extract the rows you want.
vaya putra on 11 Nov 2019 at 15:59
more complexs and its become char?
Adam Danz
on 11 Nov 2019 at 16:57
Reading in the entire file is MUCH less complex than the method you're currently using. The data are read in as char just like the method you're already using. That's why you're applying sscanf.

Sign in to comment.




2 Answers

Answer by Rik
on 12 Nov 2019 at 12:30
 Accepted Answer

I suspect this is what you need:
C=readfile('output.txt');%read the file to a cell array
Ca=C(2618:3484);%select the lines
%process every line to doubles
Cb=cellfun(@(x) sscanf(x,'%f'),Ca,'UniformOutput',false);
%convert to matrix


vaya putra on 12 Nov 2019 at 12:52
on 12 Nov 2019 at 14:31
If this solved your question, feel free to mark it as accepted answer. If not, feel free to comment with your remaining issues.

Sign in to comment.

Answer by Adam Danz
on 11 Nov 2019 at 16:57

Read in the whole file at once. In this example I'm using fileread() because I'm unfamiliar with the FEX submission readfile() mentioned in your question.
C = fileread('myTextFile.txt');
Split the char array by lines into a cell array
Ca = strsplit(C,newline);
Extract text lines n to m
If you'd rather have a char array as output


on 12 Nov 2019 at 12:34
It doesn't seem to matter for this specific file (especially for the numerical data part), but the fileread function ignores the encoding, which results in the wrong output for any char >255. Which is why I wrote this submission in the first place (and because many methods of file reading throw away blank lines).
Adam Danz
on 13 Nov 2019 at 16:03
I just had the chance to check out that function. I see why it would be useful!
on 13 Nov 2019 at 17:02
You can probably imagine the frustration that led to me writing this function. I admit it is a strange use case to want a text file reader that handles UTF8 and ANSI for both Matlab and Octave. I even managed to find a bug in the Octave implemention of textscan.
If you have any suggestions for how to make the function more robust, please feel free to comment on the FEX (or send me an email, you can get the address from the function doc).

Sign in to comment.