Finding array index for a given array element does not work all the time

Hello,

I've been using the find(ismemeber()) function to find the index of an array element, however it does not work all the time for some reason. As you can see from the picture, it returns and empty vector at some given search values. As it can be seen from the array, it should return 393 for the last line of code. I am using the latest installation of matlab at the time of writting (2023b update 7).

To me this looks like a bug with the ismemeber() function, as it returns a false for the second entry(0.3920) but not the first entry(0.3910), se last two lines of code below. However, it could just be that i am doing something wrong? As I construct the array with a start, end, and stepsize, it shouldn't be a round of error?

Best regards Anders

mf = 0:0.001:0.6;
find(ismember(mf,0.3910))

ans = 392

find(ismember(mf,0.3920))

ans =

  1×0 empty double row vector

any(ismember(mf,0.3910))

ans = logical
   1

any(ismember(mf,0.3920))

ans = logical
   0

[Stephen23]

"To me this looks like a bug with the ismemeber() function"

Nope.

"However, it could just be that i am doing something wrong?"

Yes. You are performing different operations on some binary floating point numbers and are then incorrectly assuming that these will produce exactly the same values. However that is not the case.

This is a completely expected result with binary floating point numbers:

https://www.mathworks.com/matlabcentral/answers/57444-faq-why-is-0-3-0-2-0-1-not-equal-to-zero

https://www.mathworks.com/matlabcentral/answers/316889-why-two-equal-numbers-are-not-equal

http://www.mathworks.com/matlabcentral/answers/102419-how-do-i-determine-if-the-error-in-my-answer-is-the-result-of-round-off-error-or-a-bug

http://www.mathworks.com/matlabcentral/answers/140656-about-numerical-difficulty-of-equality-comparison-operation-between-two-double-real-numbers-in-matlab

This is worth reading as well:

https://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html

https://pages.cs.wisc.edu/~david/courses/cs552/S12/handouts/goldberg-floating-point.pdf

Answer 1

[Dyuman Joshi]

Welcome to the world of floating point numbers, where not all numbers can be represented exactly in binary form.

See this thread for more information - https://in.mathworks.com/matlabcentral/answers/57444-why-is-0-3-0-2-0-1-not-equal-to-zero?s_tid=faqs_link

When comparing floating point numbers, the best practice is to use a tolerance -

mf = 0:0.001:0.6;
val = 0.3910;

tol = 1e-6;

idx1 = find(abs(mf - val) < tol)

idx1 = 392

You can also use ismembertol here -

out = ismembertol(mf,val);
idx2 = find(out)

idx2 = 392

[Anders Tagmose Lundsfryd]

Thanks for the respons, guess it was a round of error then :D

Answer 2

[Chunru]

Mind the round off error in searching double array

mf = 0:0.001:0.6;

i = find(abs(mf - 0.3910) < 1e-6)

i = 392

[Anders Tagmose Lundsfryd]

Thank you for the respons :D