Hello, I'm trying to make some calculation with vectors and matrix, with functions as sin, cos and operators, but comparing with excell...it gives me differences of 0,0004...I have tried with vpa, but...anyone can help me please?

teta_transmis(1,1)=teta+beta(1);%calculul primei orientari
if teta_transmis(1,1) > 400
    teta_transmis(1,1)=teta_transmis(1,1)-400;
end
n=length(beta);
for i=2:n
    teta_transmis(i,1)=vpa(teta_transmis(i-1,1)+200+beta(i),50);
    if teta_transmis(i,1)>400 
        teta_transmis(i,1)=vpa(teta_transmis(i,1)-400,50);
    end;
end;

[Walter Roberson]

vpa() only applies to symbolic calculations.

How different are the answers you see?

[Loghin Ana-Maria]

I have made the differences and it gives me the 0.0004 value...it remains constant... I don't know how to explain why...

[per isakson]

"I have made the differences" &nbsp How have you done that?

[Loghin Ana-Maria]

a comparison between the resulted numbers from matlab and those from excell...they are different with 0,0004

[per isakson]

Could it a result of rounding errors in the presentation?

[Loghin Ana-Maria]

I have made the same calculation in excell and the result is different from Matlab and I don't know why...

Answer 1

[John D'Errico]

This is an extremely common error made by people.

Usually this has happened not because of the computations, but because of the precision of the numbers as transferred over. For example, suppose you have a number like 1.0243 in Excel. What too many people do is to TYPE it into MATLAB as 1.0243. The problem is, Excel actually was carrying more precision than that. It was surely not the actual number. So what happens is you did the computations in MATLAB with the wrong numbers.

Why do I think this is what you have done? Because you have reported an essentially constant error, regardless of whether you used higher precision to do the computations.

[Loghin Ana-Maria]

Yes, that was the problem...thank you very much