@@ -35,15 +35,30 @@ \subsection{Koszul complex}
3535 It's indeed a complex as $ d_a^2 = 0 $ .
3636\end {definition }
3737
38- \begin {definition }[Dual Koszul complex ]
38+ \begin {definition }[Dual Koszul Complex ]
3939 \label {def:DualKoszulComplex }
40- % \leanok
40+ \leanok
41+ The chain complex dual to koszul complex via the functor $ \operatorname {Hom}(-,R)$ .
4142 \uses {def:KoszulComplex}
43+ \end {definition }
44+
45+ \begin {proposition }[Koszul Complex is Self-Dual]
46+ \label {prop:free_koszul_is_iso_to_dual }
47+ \uses {def:DualKoszulComplex}
4248 If $ M \cong R^m$ is a free $ R$ -module, then applying the contravariant functor $ \operatorname {Hom}_R(\ , R)$ gives us an isomorphic chain complex
4349 $$ K_{\bullet }(a) :\begin {tikzcd}
4450 \bigwedge ^{i+1}_R R^m \arrow [r, "d_{i+1}"] & \bigwedge ^i_R R^m \arrow [r, "d_i"] & \bigwedge ^{i-1}_R R^m \arrow [r] & \dots R^m \arrow [r, "d_1 "] & R \arrow [r] & 0
4551 \end {tikzcd}$$
46- \end {definition }
52+ \end {proposition }
53+
54+ \begin {proposition }[Differential of DualKoszul]
55+ \label {prop:diff_of_dual_koszul }
56+ \uses {def:DualKoszulComplex}
57+ % If $M \cong R^m$ is a free $R$-module, then applying the contravariant functor $\operatorname{Hom}_R(\ , R)$ gives us an isomorphic chain complex
58+ Taking $ \{ e_1 , \dots , e_m\} $ as the standard basis of $ R^m$ , we may write the differential of DualKoszul $ K(a_1 , \dots , a_m)$ as:
59+ $$ \bigwedge ^{i+1}_R R^m \to \bigwedge ^i_R R^m: \wedge \dots \wedge e_{j_{i+1}} \mapsto \sum _{k=1}^{i+1} (-1 )^{i+k} a_{j_k} e_{j_{1}} \wedge \dots \wedge \widehat {e_{j_k}} \wedge \dots \wedge e_{j_{i+1}}$$
60+ Where $ \widehat {e_{j_k}}$ means the $ j_k$ -th term is removed, so that the image is in $ \bigwedge ^{i} R^m$ .
61+ \end {proposition }
4762
4863\begin {definition }
4964 \label {def:complex_tensor_product }
@@ -54,7 +69,7 @@ \subsection{Koszul complex}
5469
5570\begin {lemma }
5671 \label {lem:koszul_of_length_2 }
57- \uses {def:DualKoszulComplex }
72+ \uses {prop:diff_of_dual_koszul }
5873 $ K(x)$ is exact exept for the rightmost side iff $ x$ is regular.
5974\end {lemma }
6075
@@ -115,7 +130,7 @@ \subsection{Koszul complex}
115130\end {theorem }
116131
117132\begin {lemma }
118- \label {lem:reg_in_jrad_iff_koszul_exact }
133+ \label {lem:reg_in_jrad_iff_koszul_exact_at_one }
119134 \uses {thm:reg_seq_implies_koszul_exact, lem:nakayama, def:reg_seq}
120135 If $ M$ is a finitely generated $ R$ -module and $ a = (a_1 , \dots , a_n)$ be a sequence in the Jacobson radical $ J$ of $ R$ .
121136 If $ H_1 (K(a) \otimes M) = 0 $ , then $ a$ is $ M$ -regular.
@@ -200,7 +215,7 @@ \subsection{Koszul complex}
200215
201216\begin {corollary }
202217 \label {cor:koszul_min_resl_residue_iff_reg }
203- \uses {lem:tor_measures_koszul_homology, cor:reg_in_jrad_iff_koszul_exact , prop:loc_resl_min_iff_basis_to_gen}
218+ \uses {lem:tor_measures_koszul_homology, lem:reg_in_jrad_iff_koszul_exact_at_one , prop:loc_resl_min_iff_basis_to_gen}
204219 If $ a_1 , \dots , a_n$ is a regular sequence in $ (R, \mathfrak {m})$ , then the Koszul complex $ K(a_1 , \dots , a_n)$ is a minimal free resolution of $ k = R/(a_1 , \dots , a_n)$ .
205220\end {corollary }
206221
0 commit comments