How to edit a particular row of a particular column of .dat file ?

I am currently working with a code which requires me edit a particular value in a column corresponding to a particular row. Following is an example of

my_dat

Some information about the file
3 !number of months

Jan   45  36  0.05
Feb   55  23  0.0016
Nov   12  7   0.45

I want to change the value 0.0016 in the last column for Feb to say 0.23. I have tried using 'textread' but only reads the first line. I also want matlab to keep the filetype after editing the file.

Answer 1

[Azzi Abdelmalek]

fid=fopen('filename.dat')
d=textscan(fid,'%s %f %f %f')
fclose(fid)
M=cell2mat(d(:,2:4))

%Or

[a,b,c,d]=textread('fic.dat',' %s %f %f %f')

[Rutvi Thanki]

Thank you for quick response. But, I get the following error

Error using cell2mat (line 45) All contents of the input cell array must be of the same data type.

[Walter Roberson]

Are you sure you used cell2mat(d(:,2:4)) instead of accidentally using cell2mat(d) ?

[Rutvi Thanki]

Yes I did use cell2mat correctly, as mentioned.

Thank you

Answer 2

[Walter Roberson]

fid = fopen('my_dat.dat', 'r');
lines{1} = fgetl(fid);
lines{2} = fgetl(fid);
lines{3} = fgetl(fid);
datacell = textscan(fid, '%s%f%f%f');
fclose(fid);

month_names = datacell{1};
month_col1 = datacell{2};
month_col2 = datacell{3};
month_col3 = datacell{4};

month_col3(2) = 0.23;

And then after that you write the data out to a new file.

There is no way to edit a text file "in place", with the exception of the uncommon case where what you are writing out is exactly the same size as what is already in the file, such as if you wrote 0.2300 to match the size of 0.0016 . Writing text files "in place" is tricky.

There are some other approaches that can be used for the case where you know the exact string you need to substitute for.

[Rutvi Thanki]

Thanks a lot for your reply.

Well I see your approach is good but this was only an example, I have a data file which has 4 columns but about 100 rows. Can you suggest a better way where I might not have to introduce 100 new variables.

Thanks