Zulip Chat Archive

Stream: Is there code for X?

Topic: Fiber at prime ideal


Wang Jingting (May 26 2025 at 07:15):

Do we have something like the following lemma?

def PrimeSpectrum.preimageOrderIsoTensorResidueField (R S : Type*) [CommRing R]
    [CommRing S] [Algebra R S] (p : PrimeSpectrum R) :
    (algebraMap R S).specComap ⁻¹' {p} o
      PrimeSpectrum (TensorProduct R p.asIdeal.ResidueField S) := sorry

I'm working on a proof_wanted on the upper bound on the Krull dimension of the polynomial ring, and this is the last sorry I'm stuck on.

Edit: This lemma corresponds to the mathematical fact that, if we have a ring homomorphism f:ABf : A \rightarrow B, let f:SpecBSpecAf^* : Spec B \rightarrow Spec A be the corresponding map between the prime spectrums, then for a prime ideal pSpecA\mathfrak{p} \in Spec A, we can identify f1pf^{*-1} \mathfrak{p} with Spec(BAκ(p))Spec (B \otimes_A \kappa ( \mathfrak{p}) ).

Andrew Yang (May 26 2025 at 08:31):

We only have the global (scheme) version. But If you only need an order iso this shouldn't be hard.
The map on the inverse direction is just TensorProduct.inclusionRight.specComap, and to show this is bijective you show that BpB_p surjects onto BAκ(p)B \otimes_A \kappa(p) and use the description of primes in localization.

Wang Jingting (May 26 2025 at 08:39):

Thanks! I'll look into this and try to add this version.


Last updated: Dec 20 2025 at 21:32 UTC