Relative primality of prime labels after localization
Let \(A\) be a commutative domain, \(\mathfrak p\) a prime ideal, and \[ \eta:A\longrightarrow A_{\mathfrak p} \] the localization map. Let \(a\in A\) be a prime element and assume \(a\nmid b\). Then \[ \eta(a),\eta(b) \] are relatively prime: every element dividing both is a unit.
More precisely, if \(a\in\mathfrak p\), then \(\eta(a)\) is prime and does not divide \(\eta(b)\). If \(a\notin\mathfrak p\), then \(\eta(a)\) is a unit. The nondivisibility assertion in the first case holds for an arbitrary commutative ring; the domain hypothesis is used to retain primality.
Relation to the OpenAI notes
Selected products and actual edge-image maps.
OpenAI notes: The notes express the outgoing kernel as the product of the supported labels in the local unique factorization domain, and extend this formula to arbitrary subsets of outgoing edges. Unsupported edges impose no condition. loc:boundary-product, loc:arbitrary-edge-subsets.
Lean (primary): The general kernel theorem assumes the DecompositionMonoid product-divisibility structure and pairwise relative primality only for the supported selected labels. Separate lower- and upper-image theorems use restriction surjectivity to identify the scalar-image submodules. Prime-label relative primality is proved after localization before specialization to the actual coset graph. PlaneLocalization:791, PlaneLocalization:795, PlaneLocalization:823, PlaneLocalization:867, PlaneLocalization:878, PlaneLocalization:918, PlaneLocalization:973.
Role in the proof: The Blueprint retains the selected-set indexing and writes both image equalities inside the same model edge module. The polynomial-domain specialization supplies every relative-primality and product-divisibility hypothesis.
Proof
Step 1. Clear a denominator in a localized divisibility
Suppose \(a\in\mathfrak p\) and, toward a contradiction, \(\eta(a)\mid\eta(b)\). Write a quotient witnessing the divisibility as \(u/t\), with \(t\notin\mathfrak p\), so \(b/1=au/t\). Equality of these localization fractions means that some \(h\notin\mathfrak p\) satisfies \[ h(tb-au)=0. \] Consequently \((ht)b=a(hu)\). The complement of a prime ideal is closed under multiplication, so \(s=ht\notin\mathfrak p\). We have obtained \[ a\mid sb \] for some \(s\notin\mathfrak p\).
Primality of \(a\) gives \(a\mid s\) or \(a\mid b\). The second contradicts the hypothesis. The first would put \(s\) in \(\mathfrak p\), since \(a\in\mathfrak p\) and an ideal contains all multiples of its elements. This also contradicts the choice of \(s\).
Thus \(\eta(a)\nmid\eta(b)\). This argument does not use cancellation in a domain.
Step 2. Retain primality when the element lies in the prime ideal
Now use that \(A\) is a domain. If \(\eta(z)=0\), the fraction criterion gives \(hz=0\) for some \(h\notin\mathfrak p\). Such \(h\) is nonzero, and cancellation gives \(z=0\). Thus the localization map is injective and \(\eta(a)\ne0\). Its image belongs to the maximal ideal of \(A_{\mathfrak p}\), hence is not a unit.
Suppose \(\eta(a)\) divides a product \((u/s)(v/t)\), where \(s,t\notin\mathfrak p\). Multiplication by the denominator units reduces this to divisibility of \(\eta(uv)\). Clearing a denominator as in Step 1 gives \[ a\mid huv \qquad\text{with }h\notin\mathfrak p. \] Primality of \(a\) and \(a\in\mathfrak p\) exclude \(a\mid h\), so \(a\mid uv\). Hence \(a\mid u\) or \(a\mid v\).
After applying \(\eta\) and dividing by the denominator units, \(\eta(a)\) divides \(u/s\) or \(v/t\). Together with nonzero and nonunit, this proves that \(\eta(a)\) is prime.
Step 3. Deduce relative primality in this case
A prime element is irreducible. If a nonunit \(d\) divided both \(\eta(a)\) and \(\eta(b)\), irreducibility of \(\eta(a)\) would make \(d\) associated to \(\eta(a)\).
Then \(\eta(a)\mid\eta(b)\), contradicting Step 1. Therefore every common divisor is a unit.
Step 4. Treat a label outside the localization prime
If \(a\notin\mathfrak p\), it is one of the inverted elements, with inverse \(1/a\). Any divisor of a unit is a unit, so \(\eta(a)\) is relatively prime to every element, including \(\eta(b)\).
The two cases prove the theorem.
For example, over a field \(k\), the two coordinate labels \(a,b\) in \(k[a,b]_{(a,b)}\) are relatively prime and both belong to the maximal ideal. Their common-divisor property is the one used to combine divisibility conditions into a product.
References
The local source declarations below give the construction and proof.
Lean formalization
OAI.KLInvariance.LocalizationRelativePrime.not_dvd_map: PlaneLocalization.lean:901.
OAI.KLInvariance.LocalizationRelativePrime.map_prime: PlaneLocalization.lean:909.
OAI.KLInvariance.LocalizationRelativePrime.map_relPrime: PlaneLocalization.lean:918.