Cartan–Eilenberg resolution

In homological algebra, the Cartan–Eilenberg resolution is in a sense, a resolution of a chain complex. It can be used to construct hyper-derived functors. It is named in honor of Henri Cartan and Samuel Eilenberg.

Definition

Let be an Abelian category with enough projectives, and let be a chain complex with objects in . Then a Cartan–Eilenberg resolution of is an upper half-plane double complex (i.e., for ) consisting of projective objects of and a chain map such that

  • Ap = 0 implies that the pth column is zero (Ppq = 0 for all q).
  • For each p, the column Pp* is a projective resolution of Ap.
  • For any fixed column,
    • the kernels of each of the horizontal maps starting at that column (which themselves form a complex) are in fact exact,
    • the same is true for the images of those maps, and
    • the same is true for the homology of those maps.

(In fact, it would suffice to require it for the kernels and homology - the case of images follows from these.) In particular, since the kernels, cokernels, and homology will all be projective, they will give a projective resolution of the kernels, cokernels, and homology of the original complex A•

There is an analogous definition using injective resolutions and cochain complexes.

The existence of Cartan–Eilenberg resolutions can be proved via the horseshoe lemma.

Hyper-derived functors

Given a right exact functor , one can define the left hyper-derived functors of F on a chain complex A∗ by constructing a Cartan–Eilenberg resolution ε : P∗∗ → A∗, applying F to P∗∗, and taking the homology of the resulting total complex.

Similarly, one can also define right hyper-derived functors for left exact functors.

gollark: Er, you need three diamonds.
gollark: Where it shines is in performing random useful tasks which there isn't dedicated hardware available for, linking together disparate systems (much more practically than redstone), working as a "microcontroller" to control something based on a bunch of input data, and entertainment-/decorative-type things (displaying stuff on monitors and whatnot, and music with Computronics).
gollark: For example, quarrying. CC has turtles. They can dig things. They can move. You can make a quarry out of this, and people have. But in practice, they're not hugely fast or efficient, and it's hard to make it work well in the face of stuff like server restarts, while a dedicated quarrying device from a mod will handle this fine and probably go faster if you can power it somehow.
gollark: I honestly don't think CC is particularly overpowered even with turtles. While it can technically do basically anything, most bigger packs will have special-purpose devices which are more expensive but do it way better, while CC is very annoying to have work.
gollark: Out of all the available APIs in _G the only ones I can see which allow I/O of some sort directly and don't just make some task you can technically already do more convenient are `fs`, `os`, `redstone`, `http`, and `term`. You can, at most, probably disable `http` and `redstone` without breaking everything horribly, and it would still be annoying.

See also

References
  • Weibel, Charles A. (1994), An introduction to homological algebra, Cambridge Studies in Advanced Mathematics, 38, Cambridge University Press, ISBN 978-0-521-55987-4, MR 1269324
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.