Read some data from a file containing numbers and characters

Hi, my problem is like this : I want to read file dynalog (.dyn) which contain text and number.

{
    **  D Y N A M I C   B E A M    S T A T I S T I C S  **

TOTAL DOSE DELIVERED             :    228 (MU)
DOSE STANDARD DEVIATION          :   0.04 (MU)
DOSE-POSITION STANDARD DEVIATION :   0.38 (deg)
NUMBER OF SAMPLES                :   1491

                     -- STT --
   INSTANCE#     DOSE    GANTRY ANGLE    (Internal scale)
                 (MU)       (deg)
       1          0.00      0.50
       2          1.45      1.50
       3          4.36      3.60
       4          7.27      5.60
       .           .          .
       .            .          .
       176        225.73    356.40
     177        227.24    358.50
     178        228.00    359.50

     -- SEGMENT BOUNDARY SAMPLES (ACTUAL) --
   INSTANCE#     DOSE    GANTRY ANGLE    (Internal scale)
                 (MU)       (deg)
       1          0.00      0.50
       2          1.47      1.23
       3          4.37      3.28
       4          7.29      5.28
       5         10.20      7.30
   ....
  ..... }

my objective is to read data after :

 -- SEGMENT BOUNDARY SAMPLES (ACTUAL) --
 INSTANCE#     DOSE    GANTRY ANGLE    (Internal scale)
               (MU)       (deg)

Any help -:)) please

[Stephen23]

Can you please upload an actual file for us to try out. Click on the paperclip button, and then both Choose file and Attach file.

[addou abdeljalil]

I have not always the same number of line. what changes in this file is the numbers control point : 1 to 178. 1 to 177 (there is not always the same number of control points)

[Stephen23]

You have a text file. If you simply change the file extension to .txt then you can upload it.

[addou abdeljalil]

ok . Thank you a lot for your help

[addou abdeljalil]

I tried the proposition of Azzi Abdelmalek it work well for filename. txt not . dyn. i do not know what to do because the machine generates DYNALOG files. I think of a conversion .text before my calculations.

[addou abdeljalil]

Can you explain me what means this character :

idx = regexpi(str,'^\s+\--(\s[\w()]+)+\s--\s*$','LineAnchors');

Thank you

Answer 1

[Azzi Abdelmalek]

p='SEGMENT BOUNDARY SAMPLES'
fid=fopen('file.txt');
a=fgetl(fid);
test=isempty(regexp(a,p));
while ischar(a) & test==1
  a=fgetl(fid)
  test=isempty(regexp(a,p));
end
fgetl(fid);
fgetl(fid);
k=1;
a=fgetl(fid);
while ischar(a);
  r{k,1}=a;
  a=fgetl(fid);
  k=k+1;
end
fclose(fid)
r

[addou abdeljalil]

I thank you Azzi. i try your proposition but i received this error : Error using fgetl (line 44) Invalid file identifier. Use fopen to generate a valid file identifier.

Error in TestReadLineFile (line 17) a=fgetl(fid);

code like this :

if true
  p = 'SEGMENT BOUNDARY SAMPLES'
fid=fopen('G:\RADIOPHY\Stages PHYS\Stage Jalil 2015\Ex BANNE\BANNE-20130801-1\2015-02-04 134243 228.dyn', 'r');
a=fgetl(fid);
test=isempty(regexp(a,p));
while ischar(a) & test==1
a=fgetl(fid)
test=isempty(regexp(a,p));
end
fgetl(fid);
fgetl(fid);
k=1;
a=fgetl(fid);
while ischar(a);
r{k,1}=a;
a=fgetl(fid);
k=k+1;
end
fclose(fid)
r
end

i think the problem comes from the way (path) of my file.

[addou abdeljalil]

When i execut my code i have

r =

    '     1          0.00      0.50'
    '     2          1.47      1.23'
    '     3          4.37      3.28'
.......
    '    89        124.12    178.63'
    '    90        125.11    180 .65'
    '    91        126.11    182.75'
    '    92        127.13    184.75'
.....

My last question is when i have 180 in the 3'th colum i will do some caclul.

if angle(i) <= 180
        a_conv(i) = abs(angle(i) - 180) ;
    else
        a_conv(i) = abs(360 - (angle(i) - 180));
    end
i tried to creat file and put *r* result in but i have some pbs. i used this code :

mat_R = fopen([CheminRes 'search180.txt'],'w'); 
if exist('search180.txt', 'file')==2
    delete('search180.txt');
end
  x = num2str(r) ;

    fprintf(mat_R,'%f\n',r);

fclose(mat_R) ;

Answer 2

[Stephen23]

Of course you can use textscan, and this is likely to be the fastest and easiest way of obtaining the file data. Here is how you can get the header data, the STT data, and the Segment Boundary Samples data:

str = fileread('temp.txt');
idx = regexpi(str,'^\s+--(\s[\w()]+)+\s--\s*$','LineAnchors');
hdr = textscan(str(1:idx(1)),'%s%f%s','HeaderLines',3,'Delimiter',{':','(',')'})
stt = textscan(str(idx(1):idx(2)),'%f%f%f','HeaderLines',4)
sbs = textscan(str(idx(2):end),'%f%f%f','HeaderLines',4)

which we can check in the command window:

hdr = 
    {4x1 cell}    [4x1 double]    {4x1 cell}
stt = 
    [7x1 double]    [7x1 double]    [7x1 double]
sbs = 
    [5x1 double]    [5x1 double]    [5x1 double]
>> hdr{1}
ans = 
    'TOTAL DOSE DELIVERED             '
    'DOSE STANDARD DEVIATION          '
    'DOSE-POSITION STANDARD DEVIATION '
    'NUMBER OF SAMPLES                '
>> stt{2}
ans =
         0
    1.4500
    4.3600
    7.2700
  225.7300
  227.2400
  228.0000
>> sbs{3}
ans =
    0.5000
    1.2300
    3.2800
    5.2800
    7.3000

Because you did not upload a sample file I used this one:

[addou abdeljalil]

Hi, I sent you my file. can you modifie your program to take account of the header of my file . your program allow me to have 2 columns that i can manipulate after easily. it is his strong point.

[Stephen23]

@addou abdeljalil: I believe this email is from you:

"Hi, I will come to you after your proposal for my matlab code.i use this code because it work well. In one line you wrote this form of code :

idx = regexpi(str,'^\s+--(\s[\w()]+)+\s--\s*$','LineAnchors');

But i do not understand all the symbols.

can you give me one explication please."

Lets have a look at the parts of this regular expression:

^                            - start of line
 \s+                         - space character, >= 1 times
    --                       - two hyphens
      (                      - start group
       \s                    - space character, once
         [\w()]+             - word or (), >= 1 times
                )+           - end group, >=1 times
                  \s+        - space character, >=1 times
                     --      - two hyphens
                       \s*   - any space characters
                          $  - end of the line

You will find the regular expressions documentation useful to read. You also might like to try using my FEX submssion RegexpHelper, which lets you interactively develop regular expressions:

http://de.mathworks.com/matlabcentral/fileexchange/48930-regular-expression-helper

[addou abdeljalil]

Thank you very much Stephen. I understood all -:))))

Answer 3

[Sid]

Can you use the Import tool ( MATLAB Doc, ) and import the data in as a cell array? You can also import headers separately, and the numeric data as a separate array as well.

Additionally, you can use textscan ( MATLAB Doc. )

[addou abdeljalil]

Thanks for your answer. i can't use textscan because i don't have same number of column. my goal is to learn my data after detecting the line quoted beolw. I would have after several files that I have to look all the time the same thing so i can't use also import data.

Other suggestion will be very appreciated.