Error using ==> mrdivide Out of memory

Question

Fadzli 2017-2-7

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/323736-error-using-mrdivide-out-of-memory

评论： Fadzli 2017-2-8

Hi, this is my code;

r=10
for i=1:r
    Cs=1.17 
    Cc=2.37 
    v=0.0000711 
    kg=0.0457 
    kp=0.26 
    Ct=2.18
    Kn=0.1 
    Cm=1.14
    T=temperature(i,:); 
    Vtx=(((2*Cs*Cc*v)*((kg/kp)+(Ct*Kn)))/((1+(3*Cm*Kn))...
        *(1+2*(kg/kp)+(2*Ct*Kn))))*(1/T)*(Delta_temperature/Delta_x); 
end

Then, there is an error comes out;

??? Error using ==> mrdivide Out of memory. Type HELP MEMORY for your options.

Error in ==> Vtx=(((2*Cs*Cc*v)*((kg/kp)+(Ct*Kn)))/((1+(3*Cm*Kn))...

What is actually happen?

p/s: Delta_temperature & Delta_x already defined...

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2017-2-7

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/323736-error-using-mrdivide-out-of-memory#answer_253672

在 MATLAB Online 中打开

You have

    T=temperature(i,:); 
    Vtx=(((2*Cs*Cc*v)*((kg/kp)+(Ct*Kn)))/((1+(3*Cm*Kn))...
        *(1+2*(kg/kp)+(2*Ct*Kn))))*(1/T)*(Delta_temperature/Delta_x);

T is a column vector, so 1/T is a request to do a Matrix Right Division. That takes more memory than a plain division -- it would require at least one matrix length(T) by length(T) to build the vandermode matrix, and then some other working memory (quite possibly another matrix the same size as that.)

Perhaps you want 1./T instead of 1/T ?

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Fadzli 2017-2-8

Thank you sir, I think I got the point...

请先登录，再进行评论。

Answer 2

Jan 2017-2-7

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/323736-error-using-mrdivide-out-of-memory#answer_253673

编辑：Jan 2017-2-7

The dimensions of the already defined variables Delta_temperature and Delta_x matter. If they are [N, 1] and [M, 1] vectors, the result is a [N, M] matrix. It seems like the creation of this matrix exhaust your memory.

There are several methods to cope with this problem:

Install more RAM
Consider if this is a bug and you want an elementwise division with the ./ operator.
Install more RAM
Close other applications
Use single precision, if applicabale and accurate enough for the problem
Install more RAM and care for using a 64 bit Matlab, such that the RAM can be used
Clear unused variables. Most of all compute large arrays outside the loop and once only
Check the sizes of the concerned variables - perhaps there is a bug before
And, you can guess it, install more RAM ;-)