The Student Room Group

group theory pairs of elements

The definition of <a,b.> I think I have from lectures is
The elements of <X> are all elements of G that can be expressed using (dotproduct) and ^(-1) from elements of X. I think X is just a subgroup of a group G.

gp4.jpggp4.jpg
(edited 8 years ago)

Scroll to see replies

Original post by cooldudeman
The definition of <a,b.> I think I have from lectures is
The elements of <X> are all elements of G that can be expressed using (dotproduct) and ^(-1) from elements of X. I think X is just a subgroup of a group G.


Agree with part (a).

Part b. I suspect you're talking about the generating set for a (sub)group. See the first two paragraphs here. Or Google further.

There is no "dotproduct", it's the group operation.

The "recall info"? Come on; what's that supposed to mean?

Those in mind, I'd reread and redo part (b). It's currently beyond redemption, IMO.
Original post by ghostwalker
Agree with part (a).

Part b. I suspect you're talking about the generating set for a (sub)group. See the first two paragraphs here. Or Google further.

There is no "dotproduct", it's the group operation.

The "recall info"? Come on; what's that supposed to mean?

Those in mind, I'd reread and redo part (b). It's currently beyond redemption, IMO.


Sorry I meant cartrsian product.

Lol I meant recall info as in the part ehere it says recall on the question.

Will have another go soon and let you know.

Thanks
Original post by cooldudeman
Sorry I meant cartrsian product.

Cartesian product doesn't come into this.


Lol I meant recall info as in the part ehere it says recall on the question.


:cool: 'See what you're saying now.

Will await your reworking.
Original post by ghostwalker
Cartesian product doesn't come into this.



:cool: 'See what you're saying now.

Will await your reworking.


Shouldn't say 34. I think I got 37 of them. I tried to think of all possibilities. Is this remotely right? If not then please help.

Posted from TSR Mobile
(edited 8 years ago)
Original post by cooldudeman
Shouldn't say 34. I think I got 37 of them. I tried to think of all possibilities. Is this remotely right? If not then please help.

Posted from TSR Mobile


I'm not sure I'm following what you're doing, but there are only 28 possible ways of choosing 2 (distinct) elements from a set of size 8, so I don't think your answer can possibly be right.
Original post by DFranklin
I'm not sure I'm following what you're doing, but there are only 28 possible ways of choosing 2 (distinct) elements from a set of size 8, so I don't think your answer can possibly be right.


Does it have to be distinct though?

Posted from TSR Mobile
Original post by cooldudeman
Does it have to be distinct though?

Posted from TSR Mobile

Well yes, since otherwise your group would be generatable from a single element and the quaternion group isn't cyclic.
Original post by cooldudeman
Shouldn't say 34. I think I got 37 of them. I tried to think of all possibilities. Is this remotely right? If not then please help.

Posted from TSR Mobile


Difficult to know how to proceed to be honest, and I have to go out in 15 minutes, with 40 minutes worth of other "stuff" to do before I do.

a,b some relationships =G\langle a,b\mid \text{ some relationships }\rangle =G

Means that the group G can be generated from the two elements a,b, subject to some relationshpis on a,b.

We'll ignore the relationship part as you're dealing with generators for a specific group whose structure is known.

Note G is a group, not just a set; there is a group operation on its elements.

So, we can reduce this to a,b=G\langle a,b\rangle =G for now; the format used in the question.

Notice also that these are equal, "<a,b>" is a group; it is not an element of G.
a,b stand for two elements of the group, AND you need to replace them with actual elements of the group, e.g <i>, just a single generator, would generate the group i^0,i^1,i^-1,i^2, etc. Which works out to 1,i,1,i{1,i,-1,-i}
So, i={1,i,i,1}\langle i \rangle=\{1,i,-i,-1\}

Now you need to find two generators, which together generate (i.e. all possible combinations of those two elements gives) the whole of Q8Q_8. Then repeat to find all possible pairs that generate the group.
(edited 8 years ago)
Original post by cooldudeman
...


Done editing.
Original post by ghostwalker
Difficult to know how to proceed to be honest, and I have to go out in 15 minutes, with 40 minutes worth of other "stuff" to do before I do.

a,b some relationships =G\langle a,b\mid \text{ some relationships }\rangle =G

Means that the group G can be generated from the two elements a,b, subject to some relationshpis on a,b.

We'll ignore the relationship part as you're dealing with generators for a specific group whose structure is known.

Note G is a group, not just a set; there is a group operation on its elements.

So, we can reduce this to a,b=G\langle a,b\rangle =G for now; the format used in the question.

Notice also that these are equal, "<a,b>" is a group; it is not an element of G.
a,b stand for two elements of the group, AND you need to replace them with actual elements of the group, e.g <i>, just a single generator, would generate the group i^0,i^1,i^-1,i^2, etc. Which works out to 1,i,1,i{1,i,-1,-i}
So, i={1,i,i,1}\langle i \rangle=\{1,i,-i,-1\}

Now you need to find two generators, which together generate (i.e. all possible combinations of those two elements gives) the whole of Q8Q_8. Then repeat to find all possible pairs that generate the group.


I'm a bit confused why you say i^-1 would give -i because is this i even the imaginary value? it doesn't say on the question, or are we supposed to assume it is. In the question it says kj=-i.

Also if we have <i, j>, we can generate

-1 fromi^2=j^2,
1 from (i^2)(j^2)=(-1)(-1)
k from ij,
-k from ji,
-i from kj=(ij)j,
-j from ik=i(ij)

Do we need to get i and j too? They are already in the <i,j> aren't they. But if we did have to 'get' them, I would do:
i from jk=j(ij)
j from ki=(ij)i

This is the whole of Q_8 right?

Now do I find the other combinations of <a,b> that also give all of Q_8?
(edited 8 years ago)
Original post by cooldudeman
I'm a bit confused why you say i^-1 would give -i because is this i even the imaginary value? it doesn't say on the question, or are we supposed to assume it is. In the question it says kj=-i.



i^0,i^1,i^-1,i^2, etc. as in i^0,i^1,i^-1,i^2,i^-2,i^3,i^-3,i^4,i^-4,....
Although it so happens i^-1 does equal -i in Q8Q_8 - try deriving it, or check it.



Also if we have <i, j>, we can generate

-1 fromi^2=j^2,
1 from (i^2)(j^2)=(-1)(-1)
k from ij,
-k from ji,
-i from kj=(ij)j,
-j from ik=i(ij)

Do we need to get i and j too?

They are already in the <i,j> aren't they. But if we did have to 'get' them, I would do:
i from jk=j(ij)
j from ki=(ij)i


You already have i,j so you don't additionally need to derive them from i,j.
Note; All combinations of i,j includes combinations of just i's, and just j's. E.g. i^2=-1, i^3=-i, etc.


This is the whole of Q_8 right?


Yep.

So, i,j=Q8\langle i,j\rangle = Q_8


Now do I find the other combinations of <a,b> that also give all of Q_8?


Yes.
(edited 8 years ago)
Original post by ghostwalker
i^0,i^1,i^-1,i^2, etc. as in i^0,i^1,i^-1,i^2,i^-2,i^3,i^-3,i^4,i^-4,....
Although it so happens i^-1 does equal -i in Q8Q_8 - try deriving it, or check it.




You already have i,j so you don't additionally need to derive them from i,j.
Note; All combinations of i,j includes combinations of just i's, and just j's. E.g. i^2=-1, i^3=-i, etc.



Yep.

So, i,j=Q8\langle i,j\rangle = Q_8



Yes.


Thanks. I feel like the only pairs we can have are <i, j> and <i, k> and <j, k>. Am I right?

Ones with -1 or +1 with i, j or k would not be able to generate X \ {i, j, k} where X is an element of {i, j, k}.

For eg, if we have <-1, i>, it cant generate j or k right?

But what would be a good reasoning to why this is???
(edited 8 years ago)
Original post by cooldudeman
Thanks. I feel like the only pairs we can have are <i, j> and <i, k> and <j, k>. Am I right?


So far, so good. There are more.


Ones with -1 or +1 with i, j or k would not be able to generate X \ {i, j, k} where X is an element of {i, j, k}.


My brain hurts reading that.


For eg, if we have <-1, i>, it cant generate j or k right?


Correct.


But what would be a good reasoning to why this is???


Well <1,i>, for example, generates the subgroup {1,-1,i,-i} - closed under the group operation (obviously, otherwise it wouldn't be a subgroup) and doesn't contain j or k, nor -j,-k
(edited 8 years ago)
Original post by ghostwalker
So far, so good. There are more.




I cant think of anymore...!

I even tried <1,-1>, it doesn't give all of it.

please help
Original post by cooldudeman
I cant think of anymore...!

I even tried <1,-1>, it doesn't give all of it.

please help


Well you've got 8 elements to choose from, and you've only really talked about 4 or 5 of them.
Original post by ghostwalker
Well you've got 8 elements to choose from, and you've only really talked about 4 or 5 of them.


oh crap, forgot about the negatives of ijk
so the only potential ones are i, j, k, -i, -j, -k so no. of ways to choose 2 from 6 elements is 6C2=15

So the only pairs that would generate Q_8 would be any combination of the elements above right?

Very tedious to write them all out, don't you think...

So it is like <x, y> with x,y in {i,j,k,-i,-j,-k} with x not equal to y, right?

EDIT: Just realised that ones like <i,-i> wont work! So that gets rid of 3 pairs so there are 12.
(edited 8 years ago)
Original post by cooldudeman
oh crap, forgot about the negatives of ijk
so the only potential ones are i, j, k, -i, -j, -k

Yes.


so no. of ways to choose 2 from 6 elements is 6C2=15

So the only pairs that would generate Q_8 would be any combination of the elements above right?


No

Does <i,-i> work?
Original post by cooldudeman

so there are 12.


Yep.
Original post by ghostwalker
Yes.



No

Does <i,-i> work?


yeah i actually edited my post. you just missed it.

So the ones would be:

i, j
-i, j
i, -j
-i, -j

i, k
-i, k
i, -k
-i, -k

j, k
-j, k
j, -k
-j, -k

right?

Quick Reply

Latest