Loop with row and column difference (multiplication)

I would like to know how I do a loop multiplying Q1 with O3 (result in Q3). Then multiply Q3 with O4 and so on for each column.

I tried to explain with other software to facilitate understanding. Thank you for your help.

[Jacqueline Rigatto]

I asked a friend that knows a little of programming to code for me, but he never used matlab

what he explained to me:

T2 is my O colun (69x1)

dir_teta is the Q, R, S, etc... (36 coluns). (1x36)

aux_T2 he created to store the result (69x36)

he came up with this code, but It's still wrong

T2=E(:,1);
nrows_2 = length(T2); 
ncols_2 = length(dir_teta);
aux_T2= ones(nrows_2,ncols_2);

for i=1:ncols_2
   aux_T2(i,1)=dir_teta(1,i);
end
for j = 1:ncols_2 
    for i = 1:nrows_2-1
    aux_T2(i+1,j)=aux_T2(i,j)*T2(i);
    end
end

[David Hill]

Did you look at my answer? First load your file and look at the data to see how it was loaded (how it is named). It would help if you attached your data file.

load('yourfile');

[Jacqueline Rigatto]

My script

Answer 1

[Cris LaPierre]

Assuming you meant to include a Q2, I'd do this using cumprod instead of a for loop.

[Jacqueline Rigatto]

Thank you for your help Cris LaPierre

Answer 2

[David Hill]

I assume your arrays when imported will start at 1. Looks like you just want cumprod() of O times initial Q.

Q(2:end)=Q(1)*cumprod(O);

[David Hill]

clear all; clc; close all

alpha=0.008;
gamma=3.3;
g=9.8; 
delta_1=0.07; 
delta_2=0.09; % if f>fp
Tp=[2 8 16];   
freq_p=1./Tp;  
i_freq=1:69;
f1=0.03093;  
c=1.1;
for i = 1:length(i_freq)
fi=f1.*c.^(i_freq-1);
end
 
freq_fi=fi;    

nrows = length(freq_fi); 
ncols = length(freq_p);
E = ones(nrows,ncols);
fc = 0.4054;  % [rad/s]


for i = 1:nrows            
    for j = 1:ncols      
        
        f = freq_fi(i);
        fp = freq_p(j);
        
        if f<fc 
            if f<=fp
            E(i,j) = (alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_1*fp))^2);
            else
            E(i,j) = (alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_2*fp))^2);
            end
        elseif f>fc
           if f<=fp
            E(i,j) = (fc/f)^5*(alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_1*fp))^2);
           else
            E(i,j) = (fc/f)^5*(alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_2*fp))^2);
           end
        else
            E(i,j)=0;
        end
    end
end
     
E;

%%%  Dir(teta)

i_teta=1:36;
teta_i=10.*i_teta;
teta_i_cos=cos(2*teta_i);

for i = 1:length(teta_i_cos)
 if teta_i_cos>0
    dir_teta=0;
else
    dir_teta= (2/pi())*(1/2)*(1+teta_i_cos);%cos^2(x) = (1+cos2x)/2
end 
end

dir_teta;

f_teta=teta_i;

%%%%%%%%%%%%% My  problem below %%%%%%%%%%%%%%%%%%%%
%%% T=2s
T2=E(:,1);
T2(T2==0)=1;%changed all T2 values that were 0 to 1 (if you multiply by zero everything below will be zero)
%when I ran it without this, the whole 69x36 matrix was zeros
aux_T2=cumprod(T2)*dir_teta;

[Jacqueline Rigatto]

aux_T2.txt came out like this, but it would be according to the T2_.txt (done in excel) .

I made a figure for better understanding:

[David Hill]

What do you do with the zero data within your table? What you are describing in red does not match your orginal question and is still confusing. Does Q4=Q1*O3*O4? That was how you described it. Or are you not placing the values into the array as you go such that Q4=orginal(Q3)*O4? What are you doing for all the white space above (between Q9 and Q12, R5 and R12, ...)?

[Jacqueline Rigatto]

Does Q4=Q1*O3*O4?

Yes.

It took me a lot of time to notice why to use cumprod(), but now I don't know how to write the code,

What I thought:

aux= cumprod(O);

being O, the Dir(alpha) array.

Now, I need do repeat this aux 69 times ( lenght of T2, that's E(f) in the figure)

and multiply each copy of aux with each value of T2

Will it work?

[Jacqueline Rigatto]

I managed to do it using the code:

clear all; clc; close all
alpha=0.008;
gamma=3.3;
g=9.8; 
delta_1=0.07; 
delta_2=0.09; % if f>fp
Tp=[2 8 16];   
freq_p=1./Tp;  
i_freq=1:69;
f1=0.03093;  
c=1.1;
for i = 1:length(i_freq)
fi=f1.*c.^(i_freq-1);
end
 
freq_fi=fi;    
nrows = length(freq_fi); 
ncols = length(freq_p);
E = ones(nrows,ncols);
fc = 0.4054;  % [rad/s]
for i = 1:nrows            
    for j = 1:ncols      
        
        f = freq_fi(i);
        fp = freq_p(j);
        
        if f<fc 
            if f<=fp
            E(i,j) = (alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_1*fp))^2);
            else
            E(i,j) = (alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_2*fp))^2);
            end
        elseif f>fc
           if f<=fp
            E(i,j) = (fc/f)^5*(alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_1*fp))^2);
           else
            E(i,j) = (fc/f)^5*(alpha*(g^2)*((2*pi)^(-4))*(f^(-5)))*exp((-5/4)*((f/fp)^(-4)))*gamma*exp((-1/2)*((f-fp)/(delta_2*fp))^2);
           end
        else
            E(i,j)=0;
        end
    end
end
     
E;
%%%  Dir(teta)
i_teta=1:36;
teta_i=10.*i_teta;
teta_i_cos=cos(2*teta_i);
for i = 1:length(teta_i_cos)
 if teta_i_cos>0
    dir_teta=0;
else
    dir_teta= (2/pi())*(1/2)*(1+teta_i_cos);%cos^2(x) = (1+cos2x)/2
end 
end
dir_teta;
f_teta=teta_i;
%%%%%%%%%%%%% My  problem below %%%%%%%%%%%%%%%%%%%%
%%% T=2s
T2=E(:,1);
T2=T2';
aux1=cumprod(dir_teta');
aux2=repmat(aux1,1,69);
aux3=aux2.*T2;

Thank you for your help David Hill

[Jacqueline Rigatto]

And I'm sorry for the confusion I made.