A general question about intersect

Hello -

I have a general question. I can provide specifics if needed.

How is it possible for

size(intersect(MyData(i),MyData)) 

to be not equal to

size(i) 

Put another way, how can the intersect of a subset of a vector and the original vector not be the entire subset of the vector? I seem to have a case where it is and I can't understand how that can be. It seems impossible, by defintion.

Thanks.

Answer 1

[Walter Roberson]

1. Duplicates in the data. An element that occurs more than once among the MyData(i) will only result in one output
2. Duplicates in the i values have the same effect.
3. NaN == NaN is never true, so the results of intersect() remove all NaN

[David McVea]

Thank you. the presence of duplicates was the reason.

Answer 2

[Image Analyst]

Because one element, MyData(i), does not necessarily match every single element of MyData. For example 2 will match only 1 element of this three element vector [1, 2, 3]. It matches only element #2 and does not match element #1 or element #3 so the length will be only 1, NOT 3.

[David McVea]

Thanks for your reply - I'm not sure this is exactly what I'm asking. from your example,

length(2) is 1, and length(intersect(2,[1 2 3]) is 1. So length(MyData(i)) equals length(i)

I think you are answering how

length(intersect(MyData(i),MyData))

can be different from

length (MyData)

I'm asking how

length(intersect(MyData(i),MyData))

can be different from

length(i)

Thanks.