matlab returns 'Error using <='

Question

Bharat 2015-6-26

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/225423-matlab-returns-error-using

评论： Walter Roberson 2015-6-26

Hello guys, I need help with a script. I have a function which contains the following commands:

    l = Y(abs(X-Sal)<0.0000001)
    p = Y(Y <= l & X>=(Sal-0.00001))

When I run, I get the following error: Error using <= Matrix dimensions must agree. However, the same script when run in the command window gives the right result. What am I doing wrong?

Thanks

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 2015-6-26

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/225423-matlab-returns-error-using#answer_184090

在 MATLAB Online 中打开

At the command window command

dbstop if error

and run the function. When it stops because of the error, look at size(Y) and size(l) . I suspect you will find that l is either empty or else has multiple elements.

What promises can you make that there will always be exactly 1 location where abs(X-Sal)<0.0000001 ?

2 个评论
显示无隐藏无

Bharat 2015-6-26

Hi Walter, Yes the l is a zero by one matrix. But I don't understand why is that happening? Regarding the 1 location thing, Sal was defined as Sal = X(3); However as abs(X=Sal) doesn't work (a problem with the number of significant digits) , I defined l as such. (I also know that the values of X do differ by atleast 0.01 which makes l exactly one value).

Walter Roberson 2015-6-26

在 MATLAB Online 中打开

You should command

format long g

and then look at the value of Sal and at the values stored in X. You should also examine the values of (X-Sal) to see what the magnitude of the difference is. If your values are above about 10^8 then the difference between consecutive double precision numbers would be more than 0.0000001; look at eps(Sal) to see the distance between adjacent numbers in that range (and realize that it is common for calculations to differ by 10*eps)

请先登录，再进行评论。