For both groups π top1 (¢ 2 − D) and π topPROOF.1 ( 2 − D) there are surjectivehomomorphisms ψ onto S n that are compatible with ı ∗ . Since δ is trivial inπ top1 ( 2 − D) we conclude that ψ(δ) = 1.By a theorem of Oka (cf. [FL, Corollary 8.4]) the short exact sequence0 → A → π top1 (¢ 2 − D)ı ∗→πtop1 ( 2 − D) → 1is central. We know that A is normally generated by δ and hence δ must be acentral element of π top (¢ 1 2 − D). Of course δ remains central in every quotient ofπ top (¢ 1 2 − D).We recall the short exact sequences1 → π top1 (X affgal ) → πtop 1 (¢ 2 − D)/ ≪ Γ i 2 ≫ ψ → S n → 1↓ ↓ ı ∗ ||1 → π top1 (X gal ) → π top1 ( 2 − D)/ ≪ Γ i 2 ≫ ψ → S n → 1We already noted that the kernel of the surjective homomorphism ı ∗ is generatedby δ. Since ψ(δ) = 1 the loop δ lies in π top1 (Xgal aff ). This yields the first exactsequence.There exist surjective homomorphismsπ top1 (Xgal aff,n−1 ) ↠ πtop 1 (X aff )↓↓π top1 (X gal , S (i)n−1) ↠ π top1 (X)The kernel of the upper horizontal homomorphism N is generated by inertiagroups. The kernel of the lower horizontal homomorphism is generated by theimage of N from above. The kernel of the left arrow downwards is generated byδ. Chasing around this diagram we find that the kernel of the surjective map fromπ top1 (X aff ) onto π top1 (X) is generated by ¯δ. □In Theorem 6.2 we constructed an isomorphismπ top1 (X affgal )/Caff ∼ =˜K(πtop1 (X aff ), n).Since δ is central it is stable under the S n -action on the right. The same holds truewhen passing to the quotient K(π top1 (X aff ), n). So if we consider K(π top1 (X aff ), n)as a subgroup of π top1 (X aff ) n then δ maps to an element of the diagonal. AndProposition 6.4 tells us exactly what this element is:72
Proposition 6.5 Under the isomorphism of Theorem 6.2 and the surjective mapof Theorem 4.7 the loop δ maps as followsπ top1 (Xgal aff)/Caff∼ top = ˜K(π 1 (Xgal aff ), n) ↠ K(πtop 1 (X aff ), n)δ ↦→ δ ↦→ (¯δ, ..., ¯δ)where ¯δ is the central element of Proposition 6.4.□Again, we can say a little bit more for the abelianisationCorollary 6.6 We keep the notations and assumptions of Theorem 6.2. Then wedenote by ¯C proj the image of C proj in the abelianised fundamental group of X gal .Then there exists an isomorphismH 1 (X gal , ¡ )/ ¯C proj ∼ = H1 (X, ¡ ) ⊕ H 1 (X aff , ¡ ) n−2 .In particular, if π top1 (X aff ) is abelian thenπ top1 (X gal )/C proj ∼ = πtop1 (X) × π top1 (X aff ) n−2 .We note that these isomorphisms are not canonical.To increase readability, we abbreviate H 1 ¡ (−, ) just by H 1 (−).Since abelianisation is not an exact functor we have to proceed by hand:PROOF.1 → 〈δ〉 → π top1 (Xgal aff)→ πtop 1 (X gal ) → 1↓ ↓ ↓〈δ ′ 〉 ↩→ H 1 (X affgal ) ↠ H 1(X gal )where δ ′ denotes the image of δ in H 1 (Xgal aff).Let x be an element of H 1(X affthat maps to 0 in H 1 (X gal ). We can lift this to an element ˜x of π top1 (Xgal aff)thathas to map to a product of commutators in π top1 (X gal ) by commutativity of thediagram. But this means that ˜x is a product of δ s for some integer s times somecommutators. Changing ˜x by commutators we still get a lift of x. So we mayassume that ˜x actually equals δ s . Therefore, x is equal to ¯δ s . This shows that wehave an exact sequence1 → 〈¯δ〉 → H 1 (X affgal ) → H 1(X gal ) → 1.We denote by ¯δ the image of δ in π top1 (X aff ) we know from Proposition 6.4 that thesubgroup generated by ¯δ inside π top1 (X aff ) is equal to the kernel of the projectionπ top1 (X aff )↠π top1 (X). So we obtain another exact sequence1 → 〈¯δ ′ 〉 → H 1 (X aff ) → H 1 (X) → 173gal )
Get now online : https://extrabook.firstmagazine.biz/?book=1491925612
D.O.W.N.L.O.A.D [P.D.F] Fundamentals of Deep Learning: Designing Next-Generation Artificial Intelligence Algorithms by Nikhil Buduma
Big Market Research presents a new package report on "Project Insight - Power Generation Construction Projects in Asia-Pacific"
Access The Complete Report On : http://www.bigmarketresearch.com/project-insight-power-generation-construction-projects-in-asia-pacific-market
The report provides detailed market analysis, information and insights based on over 1,800 projects tracked by Construction Intelligence Center (CIC). The report provides detailed metrics on each countries’ power generation projects (as tracked by CIC) split by type, start date by value, and MW capacity. The analysis is backed up by power generation capacity forecasts from GlobalData up to 2025.