Bourbaki

Remark

Tag 05KT, permanent. Cite it as [Bourbaki, Tag 05KT].

Algebra, chapter II, § 1. Modules

(2) Let E, F be two A-modules such that E is the direct sum of a finite family (Mi)1≤i≤m(M_i)_{1 \leq i \leq m} of submodules and F the direct sum of a finite family (Nj)1≤j≤n(N_j)_{1 \leq j \leq n} of submodules. Then it is known (no. 6, Corollary 1 to Proposition 6) that HomA(E,F)\mathrm{Hom}_A(E, F) is canonically identified with the product ∏i,jHomA(Mi,Nj)\prod_{i,j} \mathrm{Hom}_A(M_i, N_j); to be precise, to a family (uji)(u_{ji}), where uji∈HomA(Mi,Nj)u_{ji} \in \mathrm{Hom}_A(M_i, N_j) there corresponds the linear mapping u:E→Fu : E \to F defined as follows. It suffices to define the restriction of uu to each of the MiM_i and for each xi∈Mix_i \in M_i,

u(xi)=∑j=1nuji(xi). u(x_i) = \sum_{j=1}^n u_{ji}(x_i).
Now let G be a third A-module, the direct sum of a finite family (Pk)1≤k≤p(P_k)_{1 \leq k \leq p} of submodules; let vv be a linear mapping of F into G and let (vkj)∈∏j,kHomA(Nj,Pk)(v_{kj}) \in \prod_{j,k} \mathrm{Hom}_A(N_j, P_k) be the family corresponding to it canonically. For all xi∈Mix_i \in M_i,
v(u(xi))=∑j=1nv(uji(xi))=∑k=1p∑j=1nvkj(uji(xi)). v(u(x_i)) = \sum_{j=1}^n v(u_{ji}(x_i)) = \sum_{k=1}^p \sum_{j=1}^n v_{kj}(u_{ji}(x_i)).
Thus it is seen that if we write
wki=∑j=1nvkj∘uji∈HomA(Mi,Pk) w_{ki} = \sum_{j=1}^n v_{kj} \circ u_{ji} \in \mathrm{Hom}_A(M_i, P_k)
the family (wki)(w_{ki}) corresponds canonically to the composite linear mapping w=v∘uw = v \circ u from E to G (cf. § 10, no. 5).

Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Transcribed from the printed edition, 1998, Springer. Not translated.