alternate way , without ismember function
显示 更早的评论
Hi
I am having problems in using the ismember function, so I want to replace ismember and I need help in replacing the function.
Below is example code
INData = [
0.0120849609375
0.014190673828125
0.015838623046875
0.01214599609375
0.012481689453125
0.0123291015625
0.621002197265625
0.017181396484375
0.015167236328125
0.015350341796875
0.014556884765625
0.01739501953125
0.621002197265625
0.023162841796875
0.017578125
0.01361083984375
0.01239013671875
0.012115478515625
0.012451171875
0.01702880859375
0.01214599609375
0.012725830078125
0.0125732421875
0.012603759765625
0.012359619140625
0.014190673828125
0.013214111328125
0.013427734375
0.01531982421875
0.621002197265625
0.01239013671875
0.013092041015625
0.017791748046875
0.015960693359375
];
dMark = [7 13 30];
s_data = ismember(1:size(INData(:,1),1),dMark);
s_data =
1×34 logical array
0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0
I need help in replacing isemember function.
Thank you
采纳的回答
Star Strider
2020-10-8
1 个投票
The ismember function is doing exactly what you told it to do! What’s the problem?
20 个评论
Life is Wonderful
2020-10-8
Star Strider
2020-10-8
What do you want the function to do?
Life is Wonderful
2020-10-8
Star Strider
2020-10-8
This appears to work, and at least here, produces the same results as ismember:
s_data = sum(1:size(INData(:,1),1) == dMark(:));
and to produce a logical vector:
s_data = sum(1:size(INData(:,1),1) == dMark(:)) == 1;
With respect to the Coder, the documentation states:
C/C++ Code Generation
Generate C and C++ code using MATLAB® Coder™.
Usage notes and limitations:
- Code generation does not support cell arrays for the first or second arguments.
- Complex inputs must be single or double.
- See Code Generation for Complex Data with Zero-Valued Imaginary Parts (MATLAB Coder).
So apparently under certain dircumstances, it can be used with Coder. (I do not have the coder, so I cannot experiment with functions using it.)
Life is Wonderful
2020-10-8
Star Strider
2020-10-8
编辑：Star Strider
2020-10-8
My pleasure!
It worked when I tested it.
EDIT — (8 Oct 2020 at 13:26)
Since the first vector must be a row vector and the second a column vector, this slight change enforces that so it will work regardless of the original vector orientations:
s_data = sum(reshape((1:size(INData(:,1),1)),1,[]) == dMark(:)) == 1;
.
Life is Wonderful
2020-10-8
编辑：Life is Wonderful
2020-10-8
Star Strider
2020-10-8
That is not my code.
My code is:
s_data = sum(reshape((1:size(INData(:,1),1)),1,[]) == dMark(:)) == 1;
and it should run without error.
Life is Wonderful
2020-10-8
Life is Wonderful
2020-10-8
Star Strider
2020-10-8
My code works when I test it. I do not have the Coder, so I cannot test my code with it.
Perhaps the Coder does not recognise the (:) subscript convention that forces a column vector.
Try this, using reshape for both vectors. (Again, it runs without error and produces the correct result when I test it in R2020b.):
s_data = sum(reshape((1:size(INData(:,1),1)),1,[]) == reshape(dMark,[],1)) == 1;
.
Life is Wonderful
2020-10-8
Life is Wonderful
2020-10-8
Life is Wonderful
2020-10-8
It is working for below case
Star Strider
2020-10-8
The Coder must not be able to do that comparison. As I mentioned, I have no experience with the Coder, since I do not have it or have access to it.
I am still doing my best to see if I can come up with something the Coder likes, since it apparently also does not do implicit expansion.
Try this:
s_data = sum(bsxfun(@minus, reshape((1:size(INData(:,1),1)),1,[]), reshape(dMark,[],1)) == 0) == 1;
.
Life is Wonderful
2020-10-8
Great, Thanks a ton for efforts. You provided the right solution
It will be helpful for me if you can explain your thougths and share the explanation 
Life is Wonderful
2020-10-8
Star Strider
2020-10-8
My pleasure!
The Coder apparently does not do autoimatic explicit expansion, as MATLAB has since R2016b. In order to get the correct result, I subtracted the second argument from the first, using bsxfun, that does the expansion, so that both arguments become the same size. (There are other ways to do that, however bsxfun is the fastest and most efficient way.) The resulting matrix is 0 where the element values are the same, so the first logical operation returns a matrix where the values that are 0 as true (or 1), and using sum over the columns produces the desired vector. This creates a numeric vector that becomes a logical vector with the second logical operation, setting the elements of the vector to true if they are equal to 1. The result is the desired logical vector.
Life is Wonderful
2020-10-8
Star Strider
2020-10-8
As always, my pleasure!
This was an interesting problem!
更多回答（1 个）
Ameer Hamza
2020-10-8
An alternative approach
dMark = [7 13 30];
s_data = false(1, size(INData, 1));
s_data(dMark) = 1;
3 个评论
Life is Wonderful
2020-10-8
Ameer Hamza
2020-10-8
Don't check the output of s_data(dMark). Just run
s_data
after running my code.
Life is Wonderful
2020-10-8
类别
在 帮助中心 和 File Exchange 中查找有关 Operators and Elementary Operations 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
