Using strepp on a textfile with replacement from array.

EDIT due to comments (learning to ask better questions)

Inputs

1)I have an EXCEL array of dimensions 322x8 containing numeric data of a specific format (3.548E-07). e.g.

if true
H+  Ca  Mg  HCO3  Na  K  SO4  Cl
3.548E-07  1.183E-01  3.222E-02  1.560E-02  5.771E-01  0.000E+00  2.666E-02  8.113E-01
continued for 321 more rows
end

2)I have a text template file and contained within are 8 strings (%1 to %8) e.g.

if true
'GASES'
'*'
'SURFACE COMPLEXES'
'*'   'no mineral'  'xoh'
'SPECIES W/ Kd and DECAY        decay constant(1/s)'
'*'  0.0  0.0   0.0
'EXCHANGEABLE CATIONS'  0
'                    master     convention     ex. coef.'
'*'                     0            0       0.000e+000
'------------------------------------------------------------------------------'
'INITIAL AND BOUNDARY WATER TYPES'
1   1        !niwtype, nbwtype = number of initial and boundary waters
1    30.0                !iwtype initial, temp (C)
'        icon        guess         ctot          constrain'     ! Modern seawater
'h2o'       1        1.000d+0       1.000d+0       '  '      0  ! 
'h+'        3        %1       %1       '  '      0  !pH=8.22
'ca+2'      1        %2       %2       '  '      0
'mg+2'      1        %3       %3       '  '      0
'hco3-'     1        %4       %4       '  '      0
'na+'       1        %5       %5       '  '      0
'k+'        1        %6       %6       '  '      0
'so4-2'     1        %7       %7       '  '      0
'cl-'       1        %8       %8       '  '      0
'*'         0          0.0            0.0          '  '      0  
1    30.0                      !iwtype boundary, temp (C)  

end

Process and Output

1) I need to replace the 8 strings from the template file with the row data. I need to repeat this 322 times and save each newly created text file with exactly the same name to a directory.

Thoughts So far I have tried to use the' strrep' function failing miserably.

Please Help!!!!

[Stephen23]

We can help you achieve this. To do this you should explain the input and output requirements a bit better. We understand MATLAB quite well, so if you tell us a bit more about what you want to achieve, then we can help you with that. Showing us broken code is not so helpful, because it doesn't actually tell us what you want to happen (unless you want to fix that code, of course).

For a start: What is the size of V1 and of what data class is it? Can you please show us exactly the strings (or an example) from the text file? And finally give an example of how you want the output to look like (to be saved in the text file), or explain it a bit more.

EDIT: Thank you for editing your question, it is much nicer!

[Stephen23]

In your example template file, each of '%1', etc, occurs twice. Is this correct? Does each get replaced by the excel-data, or once only taking up the whole space between them?

[Hillaryfor2016]

Actually 4 times; full text is;

if true
      icon        guess         ctot          constrain'     ! Modern seawater
'h2o'       1        1.000d+0       1.000d+0       '  '      0  ! 
'h+'        3        %1       %1       '  '      0  !pH=8.22
'ca+2'      1        %2       %2       '  '      0
'mg+2'      1        %3       %3       '  '      0
'hco3-'     1        %4       %4       '  '      0
'na+'       1        %5       %5       '  '      0
'k+'        1        %6       %6       '  '      0
'so4-2'     1        %7       %7       '  '      0
'cl-'       1        %8       %8       '  '      0
'*'         0          0.0            0.0          '  '      0  
1    30.0                      !iwtype boundary, temp (C)  
       icon        guess         ctot          constrain'     ! Modern seawater 
'h2o'       1        1.000d+0       1.000d+0       '  '      0  ! 
'h+'        3        %1       %1       '  '      0  !pH=8.22
'ca+2'      1        %2       %2       '  '      0
'mg+2'      1        %3       %3       '  '      0
'hco3-'     1        %4       %4       '  '      0
'na+'       1        %5       %5       '  '      0
'k+'        1        %6       %6       '  '      0
'so4-2'     1        %7       %7       '  '      0
'cl-'       1        %8       %8       '  '      0
end

I just copied and pasted the first half as I thought it wouldn't matter so much.

EDIT: Each gets replaced by the excel data i.e for %1

if true
'h+'        3        3.548E-07      3.548E-07       '  '      0  !pH=8.22
....
'h+'        3        3.548E-07      3.548E-07       '  '      0  !pH=8.22
% code
end

Answer 1

[Thorsten]

A sketch how you could proceed:

1. Read the EXCEL file with xlsread

2. Create long array of char with the desired information

 text = 'if true ''GASES'' ... % up the h2o line

Append the rows with the missing data

text = [text 
sprintf(' ''h+''        3        %.3E       %.3E       ''  ''      0  !pH=8.22', 3.548E-07, 3.548E-07)]

3. write the stuff to a file using

 save(filename, text, '-ascii')

Please let me know if you got stuck somewhere when implementing along these hints.

[Hillaryfor2016]

I think this solution would work if I was to edit a singular line of the text file. Im not sure how this links to the array data, do I replace the , 3.548E-07, 3.548E-07)] with column information? i.e asssuming the array is called EXCELDATA

if true
sprintf(' ''h+''        3        %.3E       %.3E       ''  ''      0  !pH=8.22', EXCELDATA(:,1), EXCELDATA(:,1))]
end

i.e. doesnt this problem need a strrep function rather than sprint ?

[Thorsten]

CR = char(10); % carriage return

template = [...
'if true' CR...
'      icon        guess         ctot          constrain''     '...
'! Modern seawater' CR...
'''h2o''       1        1.000d+0       1.000d+0       ''  ''      0  !' CR ... 
'''h+''        3        %1       %1       ''  ''      0  !pH=8.22' CR ...
'''ca+2''      1        %2       %2       ''  ''      0' CR ...
'''mg+2''      1        %3       %3       ''  ''      0' CR ...
'''hco3-''     1        %4       %4       ''  ''      0' CR ...
'''na+''       1        %5       %5       ''  ''      0' CR ...
'''k+''        1        %6       %6       ''  ''      0' CR ...
'''so4-2''     1        %7       %7       ''  ''      0' CR ...
'''cl-''       1        %8       %8       ''  ''      0' CR ...
'''*''         0          0.0            0.0          ''  ''      0' CR ...  
'1    30.0                      !iwtype boundary, temp (C)' CR ...  
'       icon        guess         ctot          constrain''     !' CR ...
'Modern seawater ' CR ... 
'''h2o''       1        1.000d+0       1.000d+0       ''  ''      0  ! ' CR ...
'''h+''        3        %1       %1       ''  ''      0  !pH=8.22' CR ...
'''ca+2''      1        %2       %2       ''  ''      0' CR ...
'''mg+2''      1        %3       %3       ''  ''      0' CR ...
'''hco3-''     1        %4       %4       ''  ''      0' CR ...
'''na+''       1        %5       %5       ''  ''      0' CR ...
'''k+''        1        %6       %6       ''  ''      0' CR ...
'''so4-2''     1        %7       %7       ''  ''      0' CR ...
'''cl-''       1        %8       %8       ''  ''      0' CR ...
'end'];

formatstr = '%1.3E';
for i = 1:8
  template = strrep(template, sprintf('%%%d', i), formatstr);
end

data8 = [.548E-07  1.183E-01  3.222E-02  1.560E-02  ...
         5.771E-01  0.000E+00  2.666E-02  8.113E-01];

% define anonymous function to write the next command as a one-liner
flattenx = @(x) x(:);

filledtemplate = sprintf(template, ...
  repmat(flattenx(repmat(data8, [2 1])), [2 1]));

% save filled template
filenumber = 1;
filename = sprintf('myfile%d.txt', filenumber);
fid = fopen(filename, 'w');
fprintf(fid, '%s', filledtemplate);
fclose(fid);

[Hillaryfor2016]

Thankyou so much Thorsten! I edited this and got it it working!!! :)

[Hillaryfor2016]

Ahh no! I thought I had got it working but I realised that it just utilizes the first column of my excel data and misses the other 7 columns. Also it solely writes to the Sol.txt (the output file) file even if I specify a range of file numbers e.g. 1:322. Your assisstance would be most appreciated!!

if true
  % code
end
template = fileread('template.txt'); %I compared both this method and writing it out manually as a text string as in your example and they are the same.
formatstr = '%1.3E';
for i = 1:8
template = strrep(template, sprintf('%%%d', i), formatstr);
end
data = xlsread('data.xlsx');          
flattenx = @(x) x(:);
filledtemplate = sprintf(template, ...
repmat(flattenx(repmat(data, [2 1])), [2 1]));
filenumber = 1;
filename = sprintf('Sol.txt', filenumber);
fid = fopen(filename, 'w');
fprintf(fid, '%s', filledtemplate);
fclose(fid);

[Thorsten]

You have to select the column of data that you need in the filledtemplateline:

template = fileread('template.txt'); 
formatstr = '%1.3E';

% this is just to replace the %1, %2 ...%8 in the template
% with '%1.3E'; it could be done by editing he template you store, 
% but I thought it would be nicer to do it in Matlab
for i=1:8
  template = strrep(template, sprintf('%%%d', i), formatstr);
end

data = xlsread('data.xlsx');  

% this is the main loop that runs trough your data and creates the files

% I dont't know the format of data, so you have to adjust the next two lines accordingly; data_i has to be the 8 values you need to fill the template.
for i = 1:Columns_of_your_data

    data_i = data(:,i);
          
    flattenx = @(x) x(:);
    filledtemplate = sprintf(template, ...
    repmat(flattenx(repmat(data_i, [2 1])), [2 1]));

    filename = sprintf('Sol%d.txt', i);
    fid = fopen(filename, 'w');
    fprintf(fid, '%s', filledtemplate);
    fclose(fid);

end

Hope that's now clearer and you can solve your problem.

[Hillaryfor2016]

YESSSSSSSSSSSSSSSS!!!!!!! Thank you!!!!!!!!!!!!!!!!!!!!

Answer 2

[Stephen23]

Here is a pseudocode outline of how you could do this:

data = readExcel;
temp = fileread(filename);
%
xpr = cellstr(reshape(sprintf('(?<=\\s)%%%d(?=\\s)',1:8),[],8).');
%
for k = 1:rows(data)
    % extract the data for one instance:
    vec = data(k,:);
    % convert vec to cell of strings, if required.
    % vec should have one string value per cell, size 1x8.
    % replace the values in the template:
    str = regexprep(temp, xpr, vec);
    %
    fid = fopen(filename);
    fprintf(fid,str);
    fclose(fid);
end

You can use regexprep to replace all of those strings simultaneously, without any loops. Because '%1' occurs in other places in the template, I only replace instances of ' %N ', ie with a space character on both sides. You can use fileread to get the whole template as one string.

[Hillaryfor2016]

Thanks for the help Stephen, your comments have been very helpful! Wish I could accept both answers as I fudged both of the methods together to get it to work.

[Stephen23]

Glad to help. You can always click on "Votes" :)