#groups-rings-fields
406252 messages · Page 510 of 407
yes
So I'm taking
Showing its finitely generated
So I end up with this diagram
And what I want is for A^m -> I to be surjective
Snake says it suffices for A^n -> A to be surjective
Working on that now
?
no
oh
look through your book again
fine just go ahead and tell me
yeah
What was your proof?
Oh hang on a sec
I an ideal not f.g.
About that lemma
yeah
anyway it's literally almost the same
I not f.g.
exists cuz not Noeth
look at 0 -> I -> A -> A/I -> 0
A/I is f.g. if f.p. by 2.6 I is f.g.
lol
this is not a proof by contrapositive lol
I mean
this is you saying "assume not" at the start of a direct proof
Nonono
I mean
your proof is the contrapositive of
"Show for A non-noetherian there is a f.g. module not of f.p."
oh right I forgot what was actually in the book
Which is what the book said
yeah
I feel like what you proved is the
more natural statement
and I did the contrapositive of it
but according to the book haha
Also I remembered something
remember how once you thought Noeth was equivalent to acc for prime ideals?
Yes
How funny that no, it's the other condition about Noeth involving ideals for which it suffices to check on primes
Is it in AM?
Nice!
Just 2 to go before Hartshorne
yup
section 4 problems are realllly easy
since you've done AG
it's a ton of stuff about Spec and stuff and most of my proofs are like "have done b4" or "follows by correspondence thm"
Nice
oh lol
Localization is great
should I do sheaf stuff
I should probably do sheaf stuff
ugh
Sheaf stuff got easier
yeah I'll do the problems I haven't
I think u did most of the hard sheaf stuff in Hartshorne
I just want to enscheme myself

what does upside down product mean
mo2men:
$$\bigcup f$$
abs_0:
So do we use upside down sum for repeated subtraction?
Is coaddition a "commutative" comultiplication? So like the image of the map $\Delta: A \to A \otimes A$ lies in the symmetric tensors?
The_Vman:
or maybe it should be called a cocommutative comultiplication
lol
okay so
that's not exactly what cocommutative means
it's okay to have Δ(x) = a (×) b + b (×) a
Commutative is the same as m ° τ = m where τ : A (×) A -> A (×) A is the "twist" isomorphism τ(a (×) b) = b (×) a
cocommutative is τ ° Δ = Δ
Right
Isn't that what having each Δ(x) be symmetric means? I don't get your example
oh wait you're totally right lol I forgot the defintion of symmetric tensor
I was thinking it was the thing spanned by all a (×) a but I'm totally wrong
Is the condition cocommutative ever used?
I guess it would be weird if you had a commutative multiplication but non-cocommutative comultiplication. Maybe impossible for a bialgebra, I really don't know much about this stuff other than the definitions
I couldn't really say, I learned a very small slice of hopf algebra stuff for a project
you can have a hopf algebra with any combination of the properties iirc
Oh ok
so a group ring is a hopf algebra with comultiplication just g |-> g (×) g, and this is cocommutative, but taking G nonabelian we get k[G] not a commutative hopf algebra
you can get both with G commutative
That kinda makes sense
I want to say if you have an affine algebraic group G, then the coordinate ring of G will be a hopf algebra which is always commutative and is cocommutative iff G is a commutative group
maybe a (ring of functions on a) lie group would be a simpler example? It should work too
the idea is that multiplication dualizes to become comultiplication, but multiplication is just like f |-> f (×) f
I’ve never heard of cocommutative 😐
and finally sweedler's 4d hopf algebra is neither
Thanks I'll look into those examples
oh lol Wikipedia has a table with this info
In mathematics, a Hopf algebra, named after Heinz Hopf, is a structure that is simultaneously an (unital associative) algebra and a (counital coassociative) coalgebra, with these structures' compatibility making it a bialgebra, and that moreover is equipped with an antiautomor...
so basically hopf algebras are just groups but in fancier categories
that's how I think of these examples
you can categorify the definition of "hopf algebra" to a "hopf algebra object" in any (symmetric?) monoidal category
in a cartesian monoidal category (monoidal product = categorical product) this reduced down the definition of a group object, and the comultiplication Δ : G -> G×G is just the diagonal map
now if you have a functor F : C -> D which preserves monoidal products, it'll send hopf algebra objects to hopf algebra objects. Taking C to be cartesian monoidal and D = k-Vect we get that F sends group objects to hopf algebras
this generates the examples of group ring, functions on an affine algebraic group, cohomology group of a lie group, etc
neat stuff imo
Right I saw that a "hopf algebra object" in the category of sets is just a group
I'll think about the thing you said about functors and how you can create group rings, ...
well, there's a free vector space functor Set -> Vect
and F(V × W) is canonically isomorphic to F(V) (×) F(W)
so it sends group objects (groups) to hopf algebras
Yeah makes sense
that's the group ring k[G]
Oh yeah that's really nice
Suppose F/k is some field extension and F/k({αi}) is algebraic. Does {αi} contain a transcendence basis?
Okay I can Zorn up to a subset which is maximal among algebraically independent subsets of the αi. If F/k(this subset) isn't algebraic, there's some element x in F which isn't algebraic. But x satisfies a polynomial in all the αi, so there's some αi which is not algebraic over k(my subset). Adding this to my subset gives a larger one, contradiction
oh good matsumura is claiming things again and not explaining
Let p < q be primes of A = k[x1,...,xn]. Then tr.deg_k κ(q) < tr.deg_k κ(p)
why is this lol
The book says we have a surjective map k[X]/p -> k[X]/q, sure, and so the weak inequality tr.deg_k κ(q) <= tr.deg_k κ(p) is obvious
I don't see why this is obvious lol
hmm what if like
take a transcendence basis for κ(p)
wlog all the elements are actually in k[X]/p by scaling
then if you push them through the map and adjoin to k maybe κ(q) is algebraic over that?
yeah it is
every element of k[X]/q is algebraic over it by pulling back to k[X]/p
quotients of algebraic elements are algebraic so you get the whole field of fractions
groovy
are exact sequences hard?
im in that section of atiyah and im just getting fucked
and idk why is thiis shit useful
they were easy in dummit and foote wtf
and wtf there are so many theorems
am i supposed to memorize this shit?
like the theorems arent easy ilke
like an easy theorem for me would be something thats general and strong like doesnt want necessary conditions
but this shit has ' suppose this suppose this ssuppose this suppose this then this + this is that '
is this normal
what does 0 mean
Group is 0
{0}
idk thats what im asking
sends 0 to 0?
yea
should be
yea so ker of that would be 0
no not ker
img of that would be 0 which means ker is 0
which means f is injective yea
then what im sayign is right ?
yea
Yes
cool\
now wtf is this thoerem
2.9
how am i supposed to figure that out
it doesnt even make any sense
It's pretty complicated,and aluffi made it an exercise
why does it not make sense
They are all R module homomorphisms,with a special condition

Unnatural?
when you see a theorem you go yea thats obv but i wonder wahts the proof
this is not
this is just weird
yea
unnatural
totatlly unintuitive in any sense
it's pretty intuitive if you write it down and understand what it means(The statement to prove is)
Define a morphism from Hom(M*,N) to Hom(M,N) by
$a(\phi)=\phi v$ and morphism from Hom(M,N) to Hom(M',N) by $b(\phi)=\phi u$
DrunkenDrake:
Now show a and b satisfy those conditions
Yes
with u:M'-->M and v:M-->M''
now all i literally know
literally
is that ker(v) = img(u)
and img(v) = 0
?
no
nvm
M''-->0 is exact
so ker(0 map) = img(v)
wtf is ker(0map) umm
0
so img(v) is 0?
is that right
lmao okay
nvm
No,The entire set
whats the entire setr
why
Because ker of that last one is M"
You are mapping a non trivial group to a trivial group
So, Everything gets mapped to identity
yea obv
lmfao yea
okay
okay so now define phi: Hom(M'',N) --> Hom(M,N) by phi(f) = f(v)
what does f(v) mean lmfao
v is a map
composite?
okay yea composite nvm
and now defiine phi': Hom(M,N) --> Hom(M',N) by phi(f) = f(u)
basically all i want to show is ker(f(u)) = img(f(v))
right?
and we have 0 --> Hom(M,N) aswell
so ker(f(u)) must equal 0 ?
as in img(0) ?
im sorry if all of this wrong but its hard for me
right?
@solemn rain yes
i want to show ker(f(u)) = 0
@solemn rain first
(Ker phi is 0 ,which means fu=0 implies f is 0 map),is what you want to prove
Not what you stated
so i want to prove that the set of all gs such that g(f(v)) = 0 is 0
right?
thats the kernel no?
so to recap define phi: Hom(M'',N) --> Hom(M,N)
by phi(f) = f(v)
and define Phi bar : Hom(M,N) --> Hom(M',N) by phi(f) = f(u)
i know ker(phi) = img(0 map) = 0
thats a given right?
no
lmfao
i want to show that
ker(phi) = 0
= img(phi bar)
Img(phi)=ker(phi bar)
yea
= 0
wait why
what
wait
i mixed them upo didnt i
no i didnt?
yea nvm
i did
ker(phi bar) = img(phi)
okay so ker(phi bar) = { x | x(phi bar) = 0 } = {x | x(f(v))=0}
right?
lmfao im stuck
This gets confusing with words
Yes
All group homomorphisms do that
Use symbols
i am
that wont help
its just confusing
x is ah omo
it sends ech ident to iden
so f(v) must be identity
f(v) must be identity function
help im dying
okay so ker(phi bar) = { x | x(phi bar) = 0 } = {x | x(f(v))=0}
@solemn rain phi bar(x)=0 which means xu=0
@solemn rain phi bar(x)=0 which means xu=0
@carmine fossil yeda
What is f here?
yea
idk
yea i fucked that up
i fucked up kernel def
phi bar (x) = 0
which is x(u) = 0
now whats next
x must be 0
That is for phi
okay
So xu=0 implies x=0
Range of u is M"
okayyyyyyyyyy
lmfao
okay
brainlet time
now Im(phi) = { x in Hom(M,N) | x =phi(h) for some h in Hom(M'',N) }
i want to show thats 0
let me try
Actually, you have to show im(phi)=ker(phi bar)
You showed ker(phi)=0
no i fucking did not
what the fuck
i showed ker(phi bar)= 0 bro
i wanted to show ker(phi bar ) = im(phi)
and im(phi) was the kernel of the zero map
so 0
whats wrong
this is so frustrating like hell
The lack of TeX is infuriating
Same
DrunkenDrake:
okay
DrunkenDrake:
Compile Error! Click the
reaction for details. (You may edit your message)
Required to prove
Yes
Yes
and i wantt oshow img(phi bar) = ker(phi)
lmfao finally
ookay
so showing ker(phi) is 0 map
uh
what
no
yea yea i got them mixed up
cuz i have like brain damage
i want toshow ker(phi bar) = img(phi)
Yes
Use $\phi$ for representing that
DrunkenDrake:
Yes
so im(v) = M''
as im(v) = ker(0 map)
right?
so v is surjective
v is not a zero map
Yes
Now,next part
Yea
now ker(phi bar) = {x in Hom(M,N) | phi bar(x) = 0}
img(phi) = { y in Hom(M,N) | y =phi(h) for some h in Hom(M'',N) }
now let x be in ker(phi bar)
now what lmfao
Well, That's the tough part
Img(phi) is in ker(phi bar)
lmfao yea lmfao
You have to show the converse
cuz of waht i just said?
or is that trolling
@solemn rain Just trolling
The other way around
okay
If phi bar(x)=0 x=phi(h)
if phi bar (x) = 0 --> x = phi(h) for some h in Hom(M'',N)
phi bar(x) = 0 --> x(u) = 0
now u is injective thats all i know
here we go again
im stuck
fucking bullshit
Now,You use the first isomorphism theorem
yeaa thats the trick
obv
how could i have not known
how do i use first iso thoerem
okay
i got it
i think]
i should quit this shit
tysm for spending the time
hhelping me @carmine fossil

What is the book proof?
it doesntr prove it
it just says its easy
and uses first iso theorem
bnut it doesnt do it on its own
it uses the composition which idk why
wtf
Screenshot
Let $p,q$ be primes of an integral domain, with $p\subseteq q$. Then we have a natural inclusion $A_q\hookrightarrow A_p$ given by the localization map. Is it then true that $pA_p$, the maximal ideal of $A_p$ is equal to $pA_q$?
Chmonkey:
Clearly $pA_q = pA_p\cap A_q$, but do we need to do the intersection?
Chmonkey:
Actually, I think this matters not for my purposes, I bet you have to do the intersection tho TBH
I think it matters again lmfao
Atiyah Macdonald still gives me nightmares sometimes.
Why
I struggled a lot in that class. I found it insanely difficult.
I recall the first assignment being like 30 pages of writing.
Much higher workload than I was used to, and I found the content difficult. The exercises in the book are great to do, but I found DF taught me the material better.
DF carried me through AM haha
I actually still need to know if what I said is true RIP
I suspect it’s false but
It’s true when p = 0 lol
Do you still want me to look at this?
This seems very very very false
hmmm
maybe this is true
no no no take A = k[x, y] and p = (x), q = (x, y). Then pA_p contains x/y, but x/y isn't in pA_q since if it was, we'd have x/y = ax/t for not in (x, y). But we can cancel x and cross multiply to get t = ay, which is in q
@next obsidian
Rip
@upbeat burrow yea im struggling with it now
Where can I start learning abstract algebra, any websites you recommend? (I don't want to buy any books until later on)
you can check out james cook on yuotube
Ok
And get a book there
No, Libgen is not legal.
check out utube then
are you at a university?
College yeah.
check your university library, they might have a (digital) copy of an algebra book
Ok.
I anti-recommend youtube videos and websites
I think it's very hard to learn this stuff without a textbook, and especially without doing problems
Abstract Algebra by Dummit and Foote
Alright I'll check it out!
Have fun! It's a great subject
Thanks!
I'm having some trouble with the "only if part"
I don't really have any experience working with analytic functions
they are functions that are locally the limit of their taylor series
Yes, that's the extent of my knowledge
I've written up the formula for such an f, vanishing at P
I guess I could factor out (x-ai) to get a sum, but not sure how to do it in a natural way
Sorry, what's the only if part?
Showing that any function vanishing has the form sum (xi-a1)gi
if you can show it for polynomials, it's basically the same thing except that for analytic functions you add "stuff converges" every now and then
hmm I have confused myself
You can write f as a series near p and get it in that form
but why would that extend to the whole domain? Some kind of identity theorem maybe? Or I'm just being and idiot
"gi analytic functions near a1...an"
oh by "near" do they mean germs?
they mean on some open containing a1...an
ah gotcha
"analytic functions near P form a ring" I think this means the ring of germs
Hmm
I guess I can factor out (xi-ai) without sabotaging the "analyticity" because of absolute convergence or something
It just feels like I can write it in many different ways
You can
But I would expect some kind of unique decomposition
Ah
nah the decomposition isn't unique
Unique decomposition should be thought of as freeness
It's very very weird when an ideal is few
*free
I want to say you can't have a free ideal of rank > 1 but you might need noetherianness for that
if you have f(x,y) = xy you have a choice of factoring x or factoring y ; or split it up and do a mix of both
Unique decomposition should be thought of as freeness
Haven't encountered those terms yet, but I'll keep that in mind
It's the first chapter of Miles Reid's undergraduate commutative algebra
Looking forward to getting some more "machinery"
basis = linearly independent spanning set
Which is exactly the same as having a unique decomposition
Ah
but this will pretty much never happen. If I <= A is an ideal and I is free then I is principal
I really need to get a better "arbitrary ring" than Z when thinking
Yeah haha definitely
So my gotos are like
Z
Fields
polynomial rings over fields
Z/n
Quotients of polynomial rings over field (usually specific ones like k[x, y] /(y^2 - x^3)
Infinite product of copies of Z/2Z
The localization k[x]_(x) and more generally local rings
this is the ring of all rational function f/g with g(0) ≠ 0
That's some good example to play around with
yeah haha you build up a library
if I want to get really fucked up
continuous functions on the unit interval
That ring is awful
Way too big
Oh, I had never thought of that as a ring
But I can see it is
I'm still feeling very lost in this course
That sucks
Yeah, hopefully it will get better
Right now it feels like you need to find a "neat trick" for most exercises
EXACTLY
EXACTLY
EXACT
Apparently,They are pretty common
i mean they have their own fields ig
like homological alg
hopefully i am not going to see that much haha
What about Topology?
They are also apparently very common in topology
and no pictures so meh
@solemn rain yo homeboy where did you learn group stuff from
i told you
dummit and foote
homeboy
and im p bad at math so u shouldnt ask me
or maybe uir asking so u avoid those sources
yea clever
fuck groups man I don't know if I should gamble and not relearn them or just try to learn them cuz they might be important for my course
id rather just learn fields and rings better and assume groups dont exist
You don't understand rings without groups
ofc you do
Rings are literally an extension of groups
I mean I just dont remember all the sylov stuff, the izers and other hard words
which dont appear at rings as far as Ive seen
k
rings have better vibe you know im sayin
vector spaces defined on groups 😍
is it weird my course covered groups rings and a bit of fields and never mentioned modules?
You never did la,ig
modules will come later probs
I did a lot of la
Then,you dealt with lots of modules
sure, but not from like abs algebra pov
modules are generally ring theory specific
so unless ur taking a course on ring theory
modules are vector spaces on rings
you can also think of them rings acting on abelian groups
yes I know that but like ive never heard about them in algebra course
maybe in the other one Im gonna take
yea you can call them higher ig lmfao
fuck tbh I never understood how group action worked
I kinda got it very very late when I didnt need to know it anymore
Vectors are modules over a field
a group G acts on a set S
in diff eqs course
wow thank you mo2men very cool!
yeah I think it kinda clicked to me when I learned about sth called 'flow' in differnetial equation and you could kinda define it as a group action and it made a lot of sense if you wanted to imagine it
I'm just learning some module stuff too. Question: is there an abelian group which when looked at as an R-module over any ring R (and any choice of scalar multiplication by R) isn't generated by a single element?
Non cyclic Vector spaces?
Do you have a specific example though?
Take space of polynomials degree <=2
So like R^3 right?
Yes
You can take the polynomial ring R[t] and have it's action on R^3 be have constant polynomials just scale as usual and t be the linear map that shifts the coordinates down
So like t*(a, b, c) = (0, a, b)
And then R^3 under that action is generated by (1,0,0)
I think it can't be a vector space because you can use linear maps like that
Ok,Try Taking a ring and the product as a ring action on the abelian group
Like on itself?
Yeah
I don't think that's well defined as it would make R characteristic 2
Z acting on R works since R isn't a cyclic group
But I want a group on which no suitable ring and action can be found
And for R you can just take it as an R vector space
Maybe there is an equivalent statement of axiom of choice which says that is always possible?
Right so I think by assuming choice you get that every vector space has a basis, and so you can pick a big enough polynomial ring to be able to permute the basis vectors to make it "cyclic"
But I'm hoping there's a weird enough abelian group which can't be made into a vector space on which this fails
Right so I think by assuming choice you get that every vector space has a basis, and so you can pick a big enough polynomial ring to be able to permute the basis vectors to make it "cyclic"
Oh you should just be able to pick the ring of linear maps T:V to V and use that as the action, the endomorphism ring
Will your method work with a commutative R module,instead of a vector space?
I don't know enough about modules to say for sure
But I think I can reword the question.
Since an action of a ring R on an abelian group G is just an action by a subring of the group endomorphisms on G, we might as well just consider the action of the full endomorphism ring on G
So the question becomes: Is there an abelian group G so that given any element x of G there's some element y of G which is not of the form f(x) for any group homomorphism f:G to G.
do you really want y = f(x) or just y can be generated by x?
because if that's all you want you can look at G = Z/2Z x Z/3Z. No homomorphism will take (0,1) to (1,0)
but that group is generated by a single element as a module over the ring Z/2Z x Z/3Z
But you can choose the element (1,1) to generate instead
Oh ok
Because any element has finite order, and you can always find an element which has a higher order
Cool
So something like the infinite direct sum of groups Z/nZ over the naturals n should also work
Yeah that was the easy part
and then if anything lower works you get some contradiction because of coprimality stuf
ah
Actually wait
I still don't know if I'm allowed to use stuff about divisibility
it's really frustrating because i've done this stuff in the undergrad version of this class and I don't know if using that is allowed or not
Depends a bit on m lol
m is just some random maximal ideal?
Wait I guess this might be Zariski lemma
Yeah it should follow from Zariski lemma
@woven obsidian
Zariski lemma says that if K is a field that's finitely generated as a k-algebra, then K is finite over k
And finite field extension implies algebraic?
this is exactly zariaki's lemma
yeah aoi
If finite, the set {1,α,α^2,α^3,...} is linearly independent over k for any α in F
@woven obsidian
Thanks, I'll have to look into that more. We haven't really talked about field extension other than k[x]/(irreducible polynomial) at all
So for this question, my approach is that we will be able to show that Z[X] is a subring of A (just take products and sums of a). Now I have to show that it is indeed isomorphic, but I don't know how to do that.
One strategy would be like to prove that a = 10 as in 1+1+1... But that invokes the knowledge that A = Z which is true, but it's like the info isn't arriving naturally.
maybe try R=Z[X]
that would give a unique homomorphism phi : A -> Z[X] with phi(a) = X
Yes, that is what I did to prove that Z[X] is a subring of A
I don't think that makes sense
you mentioned that you constructed an inclusion map f : Z[X] -> A
I meant that there is a subring in A isomorphic to Z[X]
f is defined by mapping f(0) = 0, f(1) = 1, f(X) = a
I took the phi function, and then noted that phi(sums of a and products of a ) = sums of X and products of X which generates Z[X]. As such, there is a subring in A isomorphic to Z[X].
Now to conclude that A itself is that subring, I have to kind of show that a = sum of 1 or something.
The idea would be to use that if it is not then the homomoprhism isnt unique but i dunno how to formalise that
phi is a map from A to Z[X], if that is injective, then you can consider A a subring of Z[X]
what you're saying seems to be the reverse of that
and I don't think it's correct
I mean there is a subring in A which is isomoprhic to Z[X].
You can show that Z[X] is a subring of A by constructing an inclusion map f : Z[X] -> A
this wont involve phi
Okay thats fine too, but I still have to prove the isomorphism
This is what i ended up doing
I realized an alternative way
we can show that (Z[X],X) is also initial in the sense that there exists a unique homomorphism from it to any (R,r in R)
then you have unique homomorphisms between (A,a) and (Z[X],X) and composing them must give the identity, since the homomorphisms from (A,a) to (A,a) and (Z[X],X) to (Z[X],X) are also unique
Does anyone know cup products
@steady axle what kind of cup products? cup products in group/galois cohomology? in the cohomology of manifolds?
ah okay
if you have a specific question about cup products I could maybe give some tips
I have difficulty following use of cup product in proof of a result
I don't understand why the equality in the last line holds
The middle equality, others are fine
let me take a look
Let me know if any notation is not clear.
maybe i'm confused, but the first equality in the last line, should that be "= delta(\bar{sigma} - 1) cup ..."?
oh no wait
No
sorry I think I see now
so yeah I think the first equality is just saying that cup product respects taking delta on one side and delta^\vee^(-1) on the other
We r shuffling delta
yeah
and then the second one I think is using the previous paragraph along with the fact that when the degrees are right, the cup product pairing is jsut evaluating a homomorphism
Yeah that works without any -1 factor issue bcoz of -2 th tate cohomology grp
and then the second one I think is using the previous paragraph along with the fact that when the degrees are right, the cup product pairing is jsut evaluating a homomorphism
I had figured out this except why is cup product evaluation of homomorphism
so I think it just comes down to the tensor product map underlying this
in general, the cup product is a map
H^i(G, A) \otimes H^j(G, B) --> H^(i+j)(G, A \otimes B)
Umm for both cohomology groups of 0th order the cup product is just tensor product then we also have description of cup product in terms of cocycles but how does it help
in general, the cup product is a map
H^i(G, A) \otimes H^j(G, B) --> H^(i+j)(G, A \otimes B)
Yes with some properties relating delta morphism
if i = 0, then an element of H^0(G, A) is just an element a \in A, and a \cup b = a \otimes b
but often times when we're dealing with cup products, we might actually secretly have a map A \otimes B --> C
and we can compose with that map and then the cup product turns into
H^i(G, A) \otimes H^j(G, B) --> H^(i+j)(G, C)
in this case, A = I_G and B = Hom(I_G, Q/Z)
and there is a very natural pairing A \otimes B --> Q/Z given by a \otimes f --> f(a)
and that's what's happening here
the cup product we care about technically lives in H^(-1)(G, I_G \otimes Hom(I_G, Q/Z))
but we are just composing with the map which sends it to H^(-1)(G, Q/Z)
and the "evaluation of the homomorphism" part is coming from this tensor product
not necessarily from the cup product
you can think of the element in H^(-1)(G, I_G) as an element of (some subquotient of) I_G
and the element in H^0(G, Hom(I_G, Q/Z)) as an element of (some subquotient of) Hom(I_Q, Q_Z)
and so we can pair them I_G \otimes Hom(I_G, Q/Z) ---> Q/Z
and just get an element of Q/Z out
which is the evaluation of the homomorphism on the element
if i = 0, then an element of H^0(G, A) is just an element a \in A, and a \cup b = a \otimes b
Doesn't this also require j to be 0 too
no
H^(-1)(G,A) is a subquotient of A
meaning "a subgroup of a quotient of A"
and H^0(G,A) is "a quotient of a subgroup of A"
when you take the tate cohomology
with the little vees on top
which I haven't written
normal H^0(G,A) is a subgroup of A, and the tate H^0 is a quotient of that
normal H^(-1)(G,A) is a quotient of A, and tate H^(-1) is a subgroup of that
Yes H^-1 is ker Nm / I_G A
yes
H^(-1) is basically just H_0
which is some kind of mirrored version of H^0
which is just to say that H^(-1) behaves more like H^0 than H^1
Nice I think get it now
Doesn't this also require j to be 0 too
@steady axle I think this misconception of mine was the culprit
Ur explaination was great thanks
May I ask at what stage r u in ur math education / career
I have a phd in number theory
Awesome
and part of the title of my dissertation was "...cup products in galois cohomology"

hi shamrock hehe
yeah, for me it was "learn it once I need to know it" as opposed to like... learning it preemptively
bert, what is the context you're learning this in?
Class field theory
My first course on algebra actually did some group cohomology
The dude running it was a number theorist who really liked finite group theory and he wanted to prove schur zassenhaus
So he put enough homological algebra and group cohomology on the problem sets to do that
oh that seems like an adventure
and then nobody actually got to the result because it was too hard
This result is essential for proving that the local artin amps can be extended to abelian closure
I was 1.5 weeks from the end and gave up 😢
I did it up until like
the group cohomology
I theoretically learned why Ext is independent of resolution
I think the thing that killed me was tensor products
Like, Thomas wanted to define the bar resolution
so he used tensor products
tensor products are cruel
And part 3 of 6 on the problem was "learn about tensor products"
Lmao o boi
Tensor products are great, but it's a bit much to learn in that context
Lemme find the homework
Lol your algebra class sounded fun
Insane in what way?
what's ZG?
Z-linear combinations of elements of G
It's a ring where addition and multiplication is exactly what you'd expect
Insane in that he covered basics homological algebra/group cohomology, intro ring theory, and galois theory in 10 weeks
In a first course
Also insane as in
Challenge Problem 153 about S^|G|
was on our like week 3 hw
and was only put on there "because we hadn't defined subgroups yet"
and on hw 1 where we just learned about groups he said "prove |g| divides |G|" which is just Lagrange
homological algebra/group cohomology, intro ring theory, and galois theory
In that order? 
S^|G| was tough, i never got that
no, homological/ring theory was concurrent for weeks 0-5 and group cohomology/Galois theory was the second half
Shouldn't this say algebraically closed field? I think I am missing something
I think it should be fine? Take e.g. F(x) = x^2 + 1 in Q[x] and a = 1. Then x^2 + 1 = ((x-1)+1)^2 + 1 = (x-1)^2 + 2(x-1) + 2
@kindred mist
Like I think you can just replace all the instances xi with xi - ai and use the binomial theorem a lot
No worries I had the same reaction haha
But this is a cool result because it shows that m = (x1 - a1,...,xn - an) is the ideal of functions vanishing at the point (a1,...,an)
ok I can see how this would be useful (at least roughly speaking)
(this was an exercise from Fulton's AG pdf - the start of it of course)
Well it is useful but it's even more about like, painting the general picture of AG
Like everything starts with this correspondence between varieties and prime ideals/affine algebraic sets and radical ideals
and this is the first step towards that
The nullstellensatz is basically a generalization of this result: if k is an algebraically closed field and I some ideal of k[x1,...,xn], and we define V(I) = { (a1,...,an) in k^n : f(a1,...,an) = 0 for all f in I }, then the ideal of functions vanishing on V(I) is the radical of I
so in particular if m = (x1 - a1,...,xn - an) then V(m) = {(a1,...,an)} and so the ideal of functions vanishing at that point is r(m) = m
Nice, ok yeah I'll try prove that then (the original thing I posted that is), I was watching this video series on com. alg. and Nullstellensatz is the last video iirc, so I definitely like having stepping stones
the nullstellensatz is awesome
I found this other pdf that had like 6 different versions and one with I(V(I)) = r(I) was the strongest version, so I can try to prove my way through those weaker ones first
Haha yeah there's so so many
You should not try to prove the nullstellensatz right now
I mean
You can try
But the weak versions are still hard
And usually proven with other hard comm alg results
I am about 1/2 way through the video series on com. alg. so yeah I'm not really there yet
I remember doing NAK and tensor products and primary ideals, not much farther than that as of now though
(I am only a first year masters student though so no rush I guess)
Yeah for sure
I'm working through matsumura's book right now
Wanted to get through the first 5 sections before trying Hartshorne again
but the last problem of section 5 has been really really hard
was Hartshorne an AG or com. alg. or...?
I'm on day 3 or so?
Hartshorne is the canonical book to learn scheme theory (modern AG). I'm not saying it's good but it's what nearly everybody recommends
ah I see, I think Ponnen points to that to prep for his "Rational Points on Varieties" text
there is a seminar for this Ponnen text on my campus atm but I can't really contribute (it is like year 2 AG), but I just like to pop in it and listen
The commutative algebra seminar on the other hand I do a lot of problems for which is fun
That makes sense
Ooh sounds fun
I'm just sitting around waiting for my classes to start right now
Doing this comm alg stuff as a distraction
I won't be taking a class on anything like it until spring
Ah mine started a few weeks ago, this is why I have gotten less alg./AG done lately, also because of teaching classes
Ah yeah I decided not to TA this year because of how shit it was in spring (i.e. while online)
I'm an undergrad so it's not like, required by my program or anything
ah see I am a TA but what I didnt know until recently (like mid august) is that I am the actual teacher this semester, so thats fun
Oh wow
That's not fun to find out on short notice lol
I mean teaching is cool but not having time to prep is a little ehhhh
I was kinda suprised bc normally teaching at a university youd have a masters first
Oh really? I think at my school any PhD student can teach a course
Maybe they need to be a candidate though, I'm not sure
I just thought that was the case in general in the US
well I am a "PhD student" in the sense that I am in a fully funded program and I am expected to make it into the PhD level courses (as I plan to of course)
funny that I am like the only person without a masters in my program though
Are you in the US?
mhmm
I didnt either, maybe just something about my uni.
I don't think any of the first years at my school last year had one
My hint is to think in terms of combinatorics
Hm that's not super helpful
Okay hint 2, let m' be the number of elements satisfying g^2 = 1. Then m' = m + 1 (why?). You should try and show n - m' is even instead of showing n - m is odd
Hmm okay
no its fine
I want to try and do it myself first
I was doing it by cases
But I thought that there must be a simpler way
Yeah the proof I have in mind is pretty elegant
I think I got it
g is not equal to g^-1 whenever g^2 is not e
That's thr trick right?
yee
Yeah exactly
Thanks
$$
\begin{tikzcd}
(A\times_C B)\times_B D\arrow{d} \arrow{r} & A \times_C B\arrow{d}\arrow{r} & A\arrow{d}\
D\arrow{r} & B\arrow{r} & C
\end{tikzcd}
$$
@latent anvil
The fact that the two squares are pullbacks means the rectangle is so you get that $(A\times_C B)\times_B C = A\times_C D$
Chmonkey:
Should D be C?
No that should be D
I don't get what D is
Oh
Chmonkey:
Sorry, this
ah okay
So yeah it's a pullback
Oh I see
You do this the other way
Get the same diagram of stuff over C
Wait no
The other diagram will be stuff over B
Oh you're not doing the general case
Just the one where we want to eliminate B
I think there’s some condition that lets you do this in general but I’m not sure what it is
Luckily for projective space C = Spec Z so there’s only one way to be over it so we don’t need to worry about the maps
How do I show for an algebraically closed field k, that if f and g are coprime in k[x,y], then they are coprime in k(x)[y]?
Say you have 10 shares of TAZ with an average share price of $10. The current share price is $5. How many shares of TAZ at $5 would you need to purchase to bring the average share price down to $6. What formula would you use for this.
Taziamoma:
I need help with that
@sullen tinsel not here
Please
Read #❓how-to-get-help
They sent me here actually
Sorry are you asking Taz?
Yeah
Yes
That's why taz is here lmao
😂😂
oh lmao I see
