sparse matrix

Question

Samir 2011-8-24

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/14456-sparse-matrix

回答： Rahul Pal 2020-8-2

I am trying out to form a sparse matrix but it runs out of memory.Below is the code similar to which I have been working on A=sparse(15000,15000); inb=sparse(eye(10000,10000); it goes out of memory here.

So for forming a identity matrix (inb) I use: inb=sparse(10000,10000); for i=1:1:10000; inb(i,i)=ones; end A(1:10000,1:10000)=inb; it goes out of memory. Is there any way I can get rid of this.

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

Andrei Bobrov 2011-8-25

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/14456-sparse-matrix#answer_19669

在 MATLAB Online 中打开

blkdiag(speye(30886,30886),sparse(68709-30886,130478-30886));

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Samir 2011-8-25

Thanks Andrei.

请先登录，再进行评论。

Answer 2

the cyclist 2011-8-24

3
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/14456-sparse-matrix#answer_19624

Try the speye() command.

3 个评论
显示 1更早的评论隐藏 1更早的评论

the cyclist 2011-8-24

I have no problem running that code, and it results in the sparse arrays I expect.

Samir 2011-8-25

Thanks the cyclist it worked on the computer with high ram.

but actually my this is the real case could you suggest any better option for the following:

aeq=sparse(68709,130478);

inb=speye(30886,30886);

aeq(1:30886,1:30886)=inb;

This is running out of memory now.

I am not getting why MATLAB is not able to include the identity matrix after being sparse also.

Is there any way I can check the number of elements I can include in the array.

请先登录，再进行评论。

Answer 3

Walter Roberson 2011-8-25

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/14456-sparse-matrix#answer_19666

在 MATLAB Online 中打开

aeq=sparse(68709,130478);

generates an empty 68709 by 130478 sparse matrix, with the number of expected non-zero elements set to 0. But then you set a whole bunch of elements to non-zero, which requires reorganizing the sparse matrix as it detects more and more non-zero elements.

If you were to use

aeq=sparse([], [], [], 68709,130478, 30886);

then it would probably have less difficulty.