What is the best way to show a symbolic matrix in some tool in Guide?

Hello my friends.

I am developing a graphic interface in Guide, where I have to use matrix with variables, in this time I fill the data in a table, after of process I try to shows the final matrix in a listbox but if matrix has integer numbers, matrix is shown but if matrix has variables matlab show the next mistake.

1.- result with only integer numbers

2.- result with some variables

3.- Matlab show this error

this is my code "Resolver" Button

if true
filas = str2double(get(handles.eslabones,'String'));
filas=filas-1;
F = sym('A', [filas 4]);
B = eye(4);
C = sym('C', [4 4]);
A=0;
A=cell2mat(get(handles.uitable1,'Data'));
med = str2double(get(handles.gradrad,'String'));
if med == 2
for i=1:filas
     c1=(A(i,1))
     c2=(A(i,2))
     c3=(A(i,3))
     c4=(A(i,4))
     F(i,1)=c1;
     F(i,2)=c2;
     F(i,3)=c3;
     F(i,4)=c4;
     C=simplify([cos(F(i,4)) -sin(F(i,4))*cos(F(i,2)) sin(F(i,4))*sin(F(i,2)) F(i,1)*cos(F(i,4));
               sin(F(i,4)) cos(F(i,4))*cos(F(i,2)) -cos(F(i,4))*sin(F(i,2)) F(i,1)*sin(F(i,4));
              0 sin(F(i,2)) cos(F(i,2)) F(i,3);
              0 0 0 1]);
     eval(sprintf('A%d = C;',i));
     B=B*C;
     eval(sprintf('A%d',i))
 end
else
   for i=1:filas

       c1=(A(i,1))
       c2=(A(i,2))
       c3=(A(i,3))
       c4=(A(i,4))

       F(i,1)=c1;
       F(i,2)=c2;
       aux1=degtorad(F(i,2))
       F(i,2)=aux1
       F(i,3)=c3;
       F(i,4)=c4;
       aux1=degtorad(F(i,4))
       F(i,4)=aux1

       C=simplify([cos(F(i,4)) -sin(F(i,4))*cos(F(i,2)) sin(F(i,4))*sin(F(i,2)) F(i,1)*cos(F(i,4));
                 sin(F(i,4)) cos(F(i,4))*cos(F(i,2)) -cos(F(i,4))*sin(F(i,2)) F(i,1)*sin(F(i,4));
                0 sin(F(i,2)) cos(F(i,2)) F(i,3);
                0 0 0 1]);
       eval(sprintf('A%d = C|;',i));
       B=B*C;
       eval(sprintf('A%d',i))
   end
end
texto=[ ]; 
for k=1:4  
texto1=sprintf(' %2.0f ',B(k,:));  
texto=sprintf('%s %s \n',texto,texto1);  
end 
set(handles.resultado,'String',texto);
  end

4.- this code part is where I try to show in a list box.

if true
texto=[ ]; 
for k=1:4  
texto1=sprintf(' %2.0f ',B(k,:));  
texto=sprintf('%s %s \n',texto,texto1);  
end 
set(handles.resultado,'String',texto);
end

thanks I hope your answer.

Answer 1

[Walter Roberson]

A) Never use eval() with a symbolic expression. symbolic expressions are displayed in a language which is not MATLAB itself and is not MuPAD either, so eval() of them can produce errors or wrong results.

B) Do not use dynamic variable names http://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

C) symbolic variables are not numeric. You cannot use a %f format to display them.

D) See the code https://www.mathworks.com/matlabcentral/answers/412636-show-values-of-matriz#answer_331143

[Oscarin]

Your information has been very helpful, I'm impressed, but could you help me to organize the matrix for a better view in guide?

this was the result

that's how the code was:

if true
filas = str2double(get(handles.eslabones,'String'));
filas=filas-1;
F = sym('A', [filas 4]);
H = eye(4);
C = sym('C', [4 4]);
A=0;
A=cell2mat(get(handles.uitable1,'Data'));
med = str2double(get(handles.gradrad,'String'));
if med == 2
for i=1:filas
     c1=(A(i,1))
     c2=(A(i,2))
     c3=(A(i,3))
     c4=(A(i,4))

       F(i,1)=c1;
       F(i,2)=c2;
       F(i,3)=c3;
       F(i,4)=c4;
       C=simplify([cos(F(i,4)) -sin(F(i,4))*cos(F(i,2)) sin(F(i,4))*sin(F(i,2)) F(i,1)*cos(F(i,4));
                 sin(F(i,4)) cos(F(i,4))*cos(F(i,2)) -cos(F(i,4))*sin(F(i,2)) F(i,1)*sin(F(i,4));
                0 sin(F(i,2)) cos(F(i,2)) F(i,3);
                0 0 0 1]);
       H=H*C;
   end
else
     for i=1:filas

       c1=(A(i,1))
       c2=(A(i,2))
       c3=(A(i,3))
       c4=(A(i,4))

       F(i,1)=c1;
       F(i,2)=c2;
       aux1=degtorad(F(i,2))
       F(i,2)=aux1
       F(i,3)=c3;
       F(i,4)=c4;
       aux1=degtorad(F(i,4))
       F(i,4)=aux1

       C=simplify([cos(F(i,4)) -sin(F(i,4))*cos(F(i,2)) sin(F(i,4))*sin(F(i,2)) F(i,1)*cos(F(i,4));
                 sin(F(i,4)) cos(F(i,4))*cos(F(i,2)) -cos(F(i,4))*sin(F(i,2)) F(i,1)*sin(F(i,4));
                0 sin(F(i,2)) cos(F(i,2)) F(i,3);
                0 0 0 1]);
       H=H*C;
     end
end
fmt = repmat([repmat('%s, ', 1, size(H,2)-1), '%s\n'], 1, size(H,1));
fmt(end-1:end) = ''; 
Mc = arrayfun(@char, H, 'uniform', 0);
outstr = regexp(sprintf(fmt, Mc{:}),'\n', 'split');
set(handles.resultado,'String',outstr);
  end

[Walter Roberson]

Mc = arrayfun(@char, H, 'uniform', 0);
maxlen = max(cellfun(@length, Mc(:)));
EL1 = sprintf('%%%ds, ', maxlen);
EL2 = sprintf('%%%ds\n', maxlen);
fmt = repmat([repmat(EL1, 1, size(H,2)-1), '%s\n'], 1, size(H,1));
outstr = regexp(sprintf(fmt, Mc{:}),'\n', 'split');
set(handles.resultado, 'String', outstr, 'font', 'fixed');

Note: this will right justify each element, using the same width for each column. This code does not attempt to use the adjust the column widths to the minimum needed to align that particular column.

For example it would use

cos(a),    -5,     0
    -2,     0,     1

instead of

cos(a), -5, 0
    -2,  0, 1

[Oscarin]

thank you very much, but I have another problem when I fill in the table for example with "90" in this case are dregrees, the program only take the firt number, in the picture we can see "cos(9)" but it is "cos(90)".

and in the second case when I fill with "a1" the program separates "a" and "1" and shows a wrong result.

What can I do for the program to get "a1" and "90" as just one variable?

it's necessary to use "a1", "a2", "a3", etc. because it's a robotic analysis and each joint have "a" and if it has 3 DOF I will have "a1", "a2" and "a3", in some case "a" is "0" or integer number.

This is the code when I get data of matrix. Is a part of the code above.

if true
A=0;
A=cell2mat(get(handles.uitable1,'Data'));
end

And finally could you tell me how invert rows and columns because they show upside down and matlab shows this mistake:

Error using matlab.ui.control.UIControl/set There is no font property on the UIControl class.

Error in Cine_direc>resolver1_Callback (line 225) set(handles.resultado, 'String', outstr, 'font', 'fixed');

if true
Mc = arrayfun(@char, H, 'uniform', 0);
maxlen = max(cellfun(@length, Mc(:)));
EL1 = sprintf('%%%ds, ', maxlen);
%EL1 = sprintf('%%%ds\n', maxlen);
fmt = repmat([repmat(EL1, 1, size(H,2)-1), '%s\n'], 1, size(H,1));
outstr = regexp(sprintf(fmt, Mc{:}),'\n', 'split');
set(handles.resultado, 'String', outstr, 'font', 'fixed');
end

That's all. Thanks for all.

[Walter Roberson]

Sorry, should have been 'fontname' rather than 'font'

[Walter Roberson]

Is handles.uitable1 the same as your Tabla de parámetros ?

If it is, then it has to be using cell arrays of character vectors in order to be able to support using 'a1' in entries. We can see that the columns have ColumnFormat 'char' and are characters by the subtle fact that the columns are left justified: if they were numeric they would be right justified.

So the Data of handles.uitable1 is a cell array of character vectors. And you cell2mat() that. The cell2mat() will fail if there are multiple rows in the uitable1, unless every row just happens to have the exact same number of characters.

The result of the cell2mat() will be a char array with as many rows as there were rows in uitable1 .

Then A(i,1) will refer to the first character of the row, A(i,2) will refer to the second character of the row, and so on.

Your

A=cell2mat(get(handles.uitable1,'Data'));

should probably be

A = get(handles.uitable1,'Data');

[Walter Roberson]

Be careful about whether you are using degrees or radians. cos(90) is cosine of 90 radians. You might want to use cosd()

[Oscarin]

I have corrected the mistakes but I have one more, when I show the matrix on listbox, matrix is upside down that is to say; the first row of original matrix now is the first colum on matrix shown.

I try to understand all of code but I don't find where change for invert rows and colums.

And could you tell me where do I need to modify code to encrease space betweet columns?.

if true
Mc = arrayfun(@char, H, 'uniform', 0);
maxlen = max(cellfun(@length, Mc(:)));
EL1 = sprintf('%%%ds, ', maxlen);
%EL1 = sprintf('%%%ds\n', maxlen);
fmt = repmat([repmat(EL1, 1, size(H,2)-1), '%s\n'], 1, size(H,1));
outstr = regexp(sprintf(fmt, Mc{:}),'\n', 'split');
set(handles.resultado, 'String', outstr, 'fontname', 'fixed');
end

thanks, sorry for this but I am beginner.

[Walter Roberson]

Instead of

Mc = arrayfun(@char, H, 'uniform', 0);

try

Mc = arrayfun(@char, H, 'uniform', 0) .';

[Oscarin]

thank you very much I have solved the error and I have another question, is there some book or where can I find that parts of code?

Thanks

[Walter Roberson]

Not really, the above comes from experience. Things like sprintf() and cellfun() and arrayfun() would be in most beginning MATLAB materials.

The general technique of computing a format and using the format later is common but might not happen to be in an introductory work.

The technique of putting data elements into a cell array and expanding the cell with {:} in a sprintf() or fprintf() call gets to be a bit MATLAB (and clones) specific. In this particular case I use it only to pass in a whole bunch of character vectors because sprintf() and fprintf() do not happen to be programmed to be able to look inside a cell array. However, it can also be used in more complicated situations involving mixed data types. When you start wanting to print out multiple columns of mixed data types, your two choices are to loop row by row passing in scalars appropriate for that row -- or using the more complicated approach of building up a cell and using {:} cell expansion to build up a list of arguments. The basic use to get a cell array of character vectors formatted is practically an example you might use to illustrate the idea of cell expansion (which would be discussed in basic MATLAB references, but not necessarily in short tutorials about using MATLAB.) However, the use to build up elaborate cells for formatting multiple columns is Intermediate level rather than introductory.

regexp() is not something I would consider introductory programming. If I recall, back when I went to university, the formal study of regular expressions was part of a 4th year course on computing theory, but it might possibly have been 3rd year instead, part of the compiler construction material. I would assess regular expressions is a topic that could be introduced in first year courses, just probably not as a first programming course -- and when I say introduced, I wouldn't want to get into the theory of how it works and what could or could not be computed with regular expressions.

On the other hand, the regexp() call could be replaced with

outstr = strsplit(sprintf(fmt, Mc{:}),'\n');

as this is a very elementary use of regexp() that is basically just string matching.

All in all... if I were to rewrite the above code for tutorial purposes for beginning MATLAB use, I would probably do so by looping line by line of output. The way I wrote the code was a more advanced version written the way it was in order to avoid explicit loops.