Formatting output text file problem

I am working on producing tributary pixels input file to GAMS and I need them to be numbered in a way that distinguishes 1.1 from 1.10 and 1.100. Is there a way to print a string of characters distinguishing these differences?

To make my question clearer I have a the structure shown below. The first column is pixel number and the second has pixels flowing into this pixel. I want to print this structure into a file such that 1.1 and 1.10 are two different pixels.

1.10000000000000  0

1.20000000000000  0

1.30000000000000  0

1.40000000000000  0

1.50000000000000  0

1.60000000000000  0 1.70000000000000  0 1.80000000000000  0 1.90000000000000  0 1.10000000000000  0

2.10000000000000  [1.10000000000000,0]

2.20000000000000  [1.20000000000000,0]

2.30000000000000  [1.30000000000000,0]

2.40000000000000  [1.40000000000000,0]

2.50000000000000  [1.50000000000000,0]

2.60000000000000  [1.60000000000000,0]

2.70000000000000  [1.70000000000000,0]

2.80000000000000  [1.80000000000000,0]

2.90000000000000  [1.90000000000000,0]

2.10000000000000  [1.10000000000000,0]

3.10000000000000  [2.10000000000000,0]

. . .

10.2000000000000  [10.1000000000000,9.20000000000000,0]

10.3000000000000  [10.2000000000000,9.30000000000000,0]

10.4000000000000  [10.3000000000000,9.40000000000000,0]

10.5000000000000  [10.4000000000000,9.50000000000000,0]

10.6000000000000  [10.5000000000000,9.60000000000000,0]

10.7000000000000  [10.6000000000000,9.70000000000000,0]

10.8000000000000  [10.7000000000000,9.80000000000000,0]

10.9000000000000  [10.8000000000000,9.90000000000000,0]

10.1000000000000  [10.9000000000000,9.10000000000000,0]

I used this code:

fid = fopen('UBNTrib_mariam1.txt', 'w');

fprintf('Neighboring Tribs List\n');

fprintf('pixel,Num tribs, Tributaries');

for k=1:size(tribarray,2)

num2str(tribarray(k).pixel),num2str(tribarray(k).trib));

fprintf(fid, '%s, %s, %s\n', num2str(tribarray(k).pixel), num2str(size(tribarray(k).trib,2)));

for j = 1:size(tribarray(k).trib,2)

fprintf(fid, '%s,', num2str(tribarray(k).trib(j)));

end

fprintf(fid, '\n');

end

fclose(fid);

% view the contents of the file type UBNTrib_mariam1.txt;

'done'

However the output is: Neighboring Tribs List pixel,Num tribs, Tributaries

1.1, 1, 0,

1.2, 1, 0,

1.3, 1, 0,

. . .

10.1, 3, 10.9,9.1,0,

so the last line for instance should be : 10.10 , 3 , 10.9, 9.10,0

Thanks a lot, Mariam

[Stephen23]

You say that "I have a the structure shown below.", but it is not clearly explained how this data is stored: is this in a file, or some kind of array within MATLAB? (A structure is a specific data class in MATLAB).

It seems that the aim is to distinguish between your input data ("structure") values 1.10000000000000 and 1.10000000000000, where one should be saved as 1.1 and the other as 1.10 . But how exactly do we know which is which? They are in fact the same number.

This issue was already addressed very comprehensively by Guillaume in your earlier question: http://www.mathworks.com/matlabcentral/answers/164978

[Stephen23]

You tell us that "the last line for instance should be : 10.10 , 3 , 10.9, 9.10,0", but omit to give the raw data for this line. Supplying this might help clarify the situation a little more :)

Answer 1

[Stephen23]

Store your row and column data separately. There is no difference between 1.1 and 1.10 when it is stored as a floating point number, unless you store some kind of separate meta-data about each number... which would be more hassle than simply storing the row and column data separately.

Your easiest choice by far is to keep the row and column separate, and print then (if required) using Kelly's / Guillaume's solution to your earlier question:

fprintf('%d.%d', row, col)

Answer 2

[per isakson]

Try

    >> fprintf( '%10.8f\n', 1 )
    1.00000000

and see fprintf, Write data to text file and expand

formatSpec — Format of the output fields string

[Mariam]

This will treat both 1.1 and 1.10 similarly as well.

Both of them will be printed 1.1000000 and 1.1000000

I need the output to be 1.1 and 1.10 instead

[Stephen23]

They print the same because they are the same number.

[per isakson]

Did you try to read the documentation?

    >> fprintf( '%.1f and %.2f instead\n', 1.1, 1.1 )
    1.1 and 1.10 instead

[Stephen23]

Using the format spec will allow the number of decimal places to be specified, but it is not yet clear to me how this information is encoded in the original data: exactly how are 1.10000000000000 and 1.10000000000000 different from one another?

And even then, trying to create fprintf format strings on-the-fly is something that is best avoided... this is a great path to obfuscated code.