Timeline for Integral versus real (universal) characteristic classes
Current License: CC BY-SA 3.0
7 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Dec 17, 2013 at 17:10 | comment | added | Ricardo Andrade | Dear @Achim Krause: Yes, any free chain complex whose homology has finite type (i.e. is finitely generated in each degree) is chain equivalent to a finite type, free chain complex. See, for example, lemma 5.5.9 of Spanier's book "Algebraic topology". In fact, we can directly apply theorem 5.5.10 of Spanier's book which gives a universal coefficient theorem for cohomology in terms of cohomology (as opposed to the usual universal coefficient theorem where cohomology is expressed in terms of homology), as long as the homology has finite type. | |
Dec 17, 2013 at 16:47 | comment | added | Achim Krause | Oh, you're right. So this argument won't apply to singular cohomology, but for the spaces in question the cochain complex is quasi-isomorphic to a degreewise finitely-generated complex, so everything should work out. | |
Dec 17, 2013 at 14:42 | comment | added | Ricardo Andrade | It is essential to restrict $C$ to also be finitely generated, otherwise the first map $\operatorname{Hom}(C,\mathbb{Z}) \otimes \mathbb{R} \to \operatorname{Hom}(C,\mathbb{R})$ will not be an isomorphism in general. | |
Dec 17, 2013 at 10:10 | review | First posts | |||
Dec 17, 2013 at 10:16 | |||||
Dec 17, 2013 at 10:04 | history | undeleted | Achim Krause | ||
Dec 17, 2013 at 10:04 | history | deleted | Achim Krause | via Vote | |
Dec 17, 2013 at 9:52 | history | answered | Achim Krause | CC BY-SA 3.0 |