reading columns from a text file with errors

I have a code displayed below

    clear
  clc
  file1=input('enter file name in single quotes  ');
  fid = fopen(file1,'rt'); 
  X = textscan(fid,'%s%s%f%s%s%s%f%f%f%f%s%s%s%s',...
      'HeaderLines',7);
  fclose(fid);
  time= X{1}; status = X{6}; length=X{10};

This reads data i need from a file so that I can analyze it. The problem wiht this code is that sometimes the 4th column of data has one word of text and sometimes it has two.

the 4th column would look like this:

X{4}=
'heating'
'heating'
'heating'
'heating'
'cooling down'
etc.

When matlab reads the file it gets to the row that says 'cooling down' in the 4th column and thinks that 'down' is another column. textscan stops working at that point and I only get the data up to the point where the row that says 'cooling down' appears.

Is there any way to merge the words cooling down to say 'coolingdown' so that the code will read all of the data? Or is there another alternative? I need to read all of the data from this file. I do not care what happens to the 4th column, because I am not interested in the data from it but I need the other columns of data.

Answer 1

[Chandra Kurniawan]

Hello, I think it is realy difficult to read 'cooling down' as 'cooling_down' or 'coolingdown' using textscan. But I found new way to do that.

First you need to edit every word 'cooling down' in your text file to be 'cooling_down'

And then you can open and read the text file normally with your script.

This code bellow can be used to replace all 'cooling down' word to be 'cooling_down'

I have a text file named 'data01.txt' that contains :

time    123  0.00   heating     5:23.16  etc
time    321  0.02   heating     5:23.18  etc
time    435  0.07  cooling down 5:23.20  etc

The code:

clear;  clc;

file01 = input('enter file name : ','s');
fid01 = fopen(file01, 'rt'); 
tline = fgetl(fid01);
x = 1;
while ischar(tline) 
    a{x} = tline;
    b{x} = regexprep(a{x}, 'cooling down', 'cooling_down');
    tline = fgetl(fid01);
    x = x + 1;
end
fclose(fid01);

file02 = 'data02.txt';
fid02 = fopen(file02, 'w');
for x = 1 : length(b) 
    fprintf(fid02, '%s\n', b{x});
end
fclose(fid02);

---------------------------------

Just enter data01.txt for file01 above. And then you will get 'data02.txt' that contains :

time    123  0.00   heating     5:23.16  etc
time    321  0.02   heating     5:23.18  etc
time    435  0.07  cooling_down 5:23.20  etc

Then you can use your code to read the text file normally

clear;  clc;
file = input('enter file name in single quotes : ','s');
fid = fopen(file,'rt'); 
X = textscan(fid,'%s %f %f %s %s %s');

fclose(fid);

---------------------------------

Type X{4} in the command window and you'll get

>> X{4}

ans = 

      'heating'
      'heating'
      'cooling_down'

>>

Hope this will help you.

[Chandra Kurniawan]

Now, just edit 'data01.txt' to be :

time 123 0.00 heating 5:23.16 etc

time 321 0.02 heating 5:23.18 etc

time 435 0.07 cooling down 5:23.20 etc

time 123 0.00 heating 5:23.16 etc

time 321 0.02 heating 5:23.18 etc

time 435 0.07 cooling down 5:23.20 etc

time 435 0.07 cooling down 5:23.20 etc

time 123 0.00 heating 5:23.16 etc

And you'll get 'data02.txt' as :

time 123 0.00 heating 5:23.16 etc

time 321 0.02 heating 5:23.18 etc

time 435 0.07 cooling_down 5:23.20 etc

time 123 0.00 heating 5:23.16 etc

time 321 0.02 heating 5:23.18 etc

time 435 0.07 cooling_down 5:23.20 etc

time 435 0.07 cooling_down 5:23.20 etc

time 123 0.00 heating 5:23.16 etc

Answer 2

[Walter Roberson]

Do your columns have fixed width, or do they have a delimiter between the fields?

Answer 3

[charles atlas]

They have spaces between them. they are numeric and text data

Answer 4

[charles atlas]

In other words the data that I read in with fopen and textscan ends up looking like this:

time    123  0.00   heating     5:23.16  etc
time    321  0.02   heating     5:23.18  etc
time    435  0.07  cooling down 5:23.20  etc

The real dtat has 10 columns of numeric and text data like that but the first 5 columns basically look like that. When I say:

   X{4}=
   'heating'
   'heating'
   'heating'
IT displays it in that way and the code doesnt read 'cooling down' because it thinks 'down' is another column and the code is not programmed to read 11 columns, only 10.

[Walter Roberson]

That *is* fixed-width columns. That makes the task easier in some ways.

Unfortunately I have an appointment to go to now and cannot show appropriate code at this time.

[charles atlas]

Thank you,

allthough long, the code Chandra Kurniawan gave works fine.