How can I write a syntax in eval function?

Is it possible to write the following syntax in eval function? If yes, please write it down

I_para(i,:)=(((((sqrt(P12(i,indx)./(4.*pi))).*cos(theta(j)).*(cos(phi)).^2)) + ((sqrt(P11(i,indx)./(4.*pi))).*((sin(phi)).^2))).^2).* (pi.*(r_med(i)).^2.*Qsca_c(i));

[John D'Errico]

You want to use eval on that? Please don't. There can be no good reason to do so.

[Wiqas Ahmad]

Using eval command, I can better interpret my program output and will be understood that's why. I tried some syntax to convert but shows the following error.

eval(['I_para',num2str(i),num2str(j),'=','(','(sqrt(P12',num2str(i),'./(4.*pi)',').*cos(theta',num2str(j),'.*(cos(phi)).^2)','+','(sqrt(P11',num2str(i),'./(4.*pi)','.*(sin(phi)).^2)',').^2','.* (pi.*(r_med',num2str(i),'.^2.*Qsca_c',num2str(i),')',';']);

Error using polar_2 (line 25)

Error: Invalid expression. When calling a function or indexing a variable, use parentheses. Otherwise, check for mismatched

delimiters.

[Rik]

Why do you want to replace indexing with obfuscated code?

The real solution is to turn your P variables into an array as well. That way you can index both.

[Stephen23]

"Using eval command, I can better interpret my program output and will be understood that's why."

In reality it will make it much more complex (as the mere existence of this question should be warning you).

"I tried some syntax to convert but shows the following error."

Perhaps the difficulty of debugging one simple line of code gives you one clue why that approach should be avoided:

https://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

Using EVAL forces you into writing slow, complex, inefficient, buggy code that is hard to debug.

The simple and efficient approach is to use indexing, just like MATLAB is designed for.

[Wiqas Ahmad]

clc;
clear all;
close all;
%% ------------------------------Program-------------------------------------
N0=100;
r_med=[0.445 0.889 1.445 8];
sigma_g=7;
N_ang=91;
theta=[0 20]/180*pi;
t = [0:180]/180*pi;
phi =[0 30]/180*pi;
Num_r = 50e3;
r = linspace(1,50,Num_r)./2;
col=['k' 'b' 'r' 'g'];
for i=1:length(r_med)
    hold on
    for j=length(theta)
        [P11(i,:),P12(i,:),P33(i,:),P34(i,:),Qsca_c(i,:),~,~] = ZK_W_Cloud_PhaseFunc(N0,r_med(i),sigma_g,N_ang);
        [~,indx] = find(theta(j)==t);
        eval(['I_para',num2str(i),num2str(j),'indx','=','(','(sqrt(P12',num2str(i),'indx','./(4.*pi)',').*cos(theta',num2str(j),'.*(cos(phi)).^2)','+(','(sqrt(P11',num2str(i),'./(4.*pi)','.*(sin(phi)).^2)',').^2','.*(','pi.*(r_med',num2str(i),'.^2.*Qsca_c',num2str(i),')',')',';']);
        eval(['I_perp',num2str(i),num2str(j),'indx','=','(','(sqrt(P12',num2str(i),'indx','./(4.*pi)',').*cos(theta',num2str(j),'.*(cos(phi)).^2)','-(','(sqrt(P11',num2str(i),'./(4.*pi)','.*(sin(phi)).^2)',').^2','.*(','pi.*(r_med',num2str(i),'.^2.*Qsca_c',num2str(i),')',')',';']);
        %I_para(i,:)=(((((sqrt(P12(i,indx)./(4.*pi))).*cos(theta(j)).*(cos(phi)).^2)) + ((sqrt(P11(i,indx)./(4.*pi))).*((sin(phi)).^2))).^2).* (pi.*(r_med(i)).^2.*Qsca_c(i));
        %I_perp(i,:)=((((sqrt(P12(i,indx)./(4.*pi))).*cos(theta(j)).*cos(phi).*sin(phi))-((sqrt(P11(i,indx)./(4.*pi))).*sin(phi).*(cos(phi)))).^2).*(pi.*(r_med(i)).^2.*Qsca_c(i));
        p1=eval(['plot(I_para',num2str(i),num2str(j),',phi)',';']);
        p1=eval(['plot(I_perp',num2str(i),num2str(j),',phi)',';']);
        p1.LineWidth=1;
        %plot(phi,I_para(i,j),'color',col(i),'Linewidth',1.5);
        %plot(phi.*180/pi,I_perp(i,j),'color',col(i),'Linewidth',1.5);
     end
end
%%
figure(1)
subplot(1,2,1);
hold on
plot(phi,I_para11,'color',col(1),'Linewidth',1.5);
plot(phi,I_para21,'color',col(2),'Linewidth',1.5);
plot(phi,I_para31,'color',col(3),'Linewidth',1.5);
plot(phi,I_para41,'color',col(4),'Linewidth',1.5);
%title('\bf (a) Scattering phase function','FontSize',...
%16,'FontWeight','normal')%\bf(bold font),\rm(normal font),\it(italian font)
leg=legend('R_{eff}= 4\mum','R_{eff}= 8\mum','R_{eff}= 13\mum','R_{eff}= 18\mum','location','Northeast',...
    'orientation','vertical','Fontsize',6.5);% we need to calculate Reff from Rm using formaula
legend boxon
leg.ItemTokenSize = [15,18];
xlabel('\phi(\circ)');
ylabel('I_{||}');
set(gca,'color','w','Fontsize',12,'LineWidth',1,'Fontweight','normal');
set(gca,'box','on','Fontname','Arial','Fontsmoothing','on');
set(gca,'xlim',[0 90],'xtick',[0:30:90],'ylim',[0 6].*10.^-2,'ytick',[0:2:6].*10.^-2);
set(gca,'xgrid','on','ygrid','on','gridcolor','k');
%%
subplot(1,2,2);
hold on
plot(phi,I_perp11,'color',col(1),'Linewidth',1.5);
plot(phi,I_perp21,'color',col(2),'Linewidth',1.5);
plot(phi,I_perp31,'color',col(3),'Linewidth',1.5);
plot(phi,I_perp41,'color',col(4),'Linewidth',1.5);
%title('\bf (a) Scattering phase function','FontSize',...
%16,'FontWeight','normal')%\bf(bold font),\rm(normal font),\it(italian font)
leg=legend('R_{eff}= 4\mum','R_{eff}= 8\mum','R_{eff}= 13\mum','R_{eff}= 18\mum','location','Northeast',...
    'orientation','vertical','Fontsize',6.5);% we need to calculate Reff from Rm using formaula
legend boxon
leg.ItemTokenSize = [15,18];
xlabel('\phi(\circ)');
ylabel('I_{\perp}');
set(gca,'color','w','Fontsize',12,'LineWidth',1,'Fontweight','normal');
set(gca,'box','on','Fontname','Arial','Fontsmoothing','on');
set(gca,'xlim',[0 90],'xtick',[0:30:90],'ylim',[0 12].*10.^-3,'ytick',[0:3:12].*10.^-3);
set(gca,'xgrid','on','ygrid','on','gridcolor','k');
%%
% Enlarge figure to full screen.
set(gcf, 'units','normalized','outerposition',[.3 .3 .45 .38]);
%print(gcf,'figure.tiff','-dtiff','-r300');

This is my whole program. I want to change the index i while keep the index j constant in the plot command (plot(phi,I_para11,'color',col(1),'Linewidth',1.5);) . Previously, I had done the similar plotting using eval commnad. Now I'm trying to use it again but the structure of my program is quite different and I can't use it now. I tried but facing some errors i don't know.@Rik

[Sargondjani]

Others were suggesting to use indexing. For example, with a cell array:

I_para{1,1} = ....

I_para{1,2} = ...

[DGM]

The error message means exactly what it says. The expressions are invalid. This is regardless of the abuse of eval(). They're invalid expressions in any context. I can't reasonably guess which parentheses are missing. You'll have to figure out what the expression is supposed to be.

i = 11; % placeholders
j = 22;

exp1 = ['I_para',num2str(i),num2str(j),'=','(','(sqrt(P12',num2str(i),'./(4.*pi)',').*cos(theta',num2str(j),'.*(cos(phi)).^2)','+','(sqrt(P11',num2str(i),'./(4.*pi)','.*(sin(phi)).^2)',').^2','.* (pi.*(r_med',num2str(i),'.^2.*Qsca_c',num2str(i),')',';']

exp1 = 'I_para1122=((sqrt(P1211./(4.*pi)).*cos(theta22.*(cos(phi)).^2)+(sqrt(P1111./(4.*pi).*(sin(phi)).^2)).^2.* (pi.*(r_med11.^2.*Qsca_c11);'

[nnz(exp1 == '(') nnz(exp1 == ')')]

ans = 1×2

14 11

exp2 = ['I_para',num2str(i),num2str(j),'indx','=','(','(sqrt(P12',num2str(i),'indx','./(4.*pi)',').*cos(theta',num2str(j),'.*(cos(phi)).^2)','+(','(sqrt(P11',num2str(i),'./(4.*pi)','.*(sin(phi)).^2)',').^2','.*(','pi.*(r_med',num2str(i),'.^2.*Qsca_c',num2str(i),')',')',';']

exp2 = 'I_para1122indx=((sqrt(P1211indx./(4.*pi)).*cos(theta22.*(cos(phi)).^2)+((sqrt(P1111./(4.*pi).*(sin(phi)).^2)).^2.*(pi.*(r_med11.^2.*Qsca_c11));'

[nnz(exp2 == '(') nnz(exp2 == ')')]

ans = 1×2

15 12

exp3 = ['I_perp',num2str(i),num2str(j),'indx','=','(','(sqrt(P12',num2str(i),'indx','./(4.*pi)',').*cos(theta',num2str(j),'.*(cos(phi)).^2)','-(','(sqrt(P11',num2str(i),'./(4.*pi)','.*(sin(phi)).^2)',').^2','.*(','pi.*(r_med',num2str(i),'.^2.*Qsca_c',num2str(i),')',')',';']

exp3 = 'I_perp1122indx=((sqrt(P1211indx./(4.*pi)).*cos(theta22.*(cos(phi)).^2)-((sqrt(P1111./(4.*pi).*(sin(phi)).^2)).^2.*(pi.*(r_med11.^2.*Qsca_c11));'

[nnz(exp3 == '(') nnz(exp3 == ')')]

ans = 1×2

15 12

Is there any reason why the constant portions of the expression are split into randomly-sized chunks?

['indx','=','(','(sqrt(P12']

instead of just

['indx=((sqrt(P12']

This seems unnecessary and makes everything more difficult to read.

Mistakes are an eventuality. Unnecessarily complicating things is basically asking for those mistakes to be obfuscated the moment they're made. This is a core element of the warnings already given about eval().

[Stephen23]

As DGM explains, by using EVAL you force yourself into writing unnecessarily complex code. But the disadvantages are not only that your code is more complex and obfuscated (than if you used basic indexing), but that you also remove all of MATLAB's inbuilt code helping tools (e.g. mlint static code checking, syntax error highlighting, variable highlighting, code warnings, etc.) which would help you to debug errors. So your design decision makes your code harder to debug. Your question is a good example of this.

Dynamically accessing variable names is also slow and inefficient compared to using basic MATLAB indexing.

https://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval