how to find index of a cell that contains a numeric value?

Question

Terek Li 2016-10-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/307666-how-to-find-index-of-a-cell-that-contains-a-numeric-value

评论： Adam 2016-10-17

Hi, there are several posts regarding this question, but none of their solutions actually work...

I have a 1000*1 cell containing random double values, how do I find the index of the cell that contains a specific value x?

I tried to do this:

index = find([data(:,1)] == x)

but I get an error saying: Undefined operator '==' for input arguments of type 'cell'.

Please help, thank you!

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Guillaume 2016-10-17

Any reason you're using a cell array for storing scalar values? Using a matrix would be a lot simpler since the above code would then work (and matrices are a lot more memory efficient).

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Adam 2016-10-17

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/307666-how-to-find-index-of-a-cell-that-contains-a-numeric-value#answer_239362

在 MATLAB Online 中打开

find( cellfun( @(d) isequal( d, x ), data) );

works theoretically, but you should never be testing for exact equality with doubles if they are not integer-valued. You should instead replace isequal(...) with some function of your own that uses a tolerance for comparison.

2 个评论
显示无隐藏无

Terek Li 2016-10-17

why? is finding doubles inefficient?

Adam 2016-10-17

It's not a matter of efficiency it is a matter of the level of accuracy with which a value can be represented in a double. Floating point values cannot be 100% accurately represented so what should be the same value, being arrived at by two different mathematical processes can be represented slightly differently - e.g. 1e-10 different from each other, something that as far as makes no difference is equal to us as humans, but will fail an equality test.

You can read up on this anywhere on the internet related to floating point accuracy if you want.

请先登录，再进行评论。