I am having a hard moment trying to solve something that looks really easy but I cannot solve. I want to prove that, given two presentations of left R-modules, one injective and the other projective, respectively, $0\to A \to I \to B \to 0$ and $0\to C \to P \to D \to 0,$ then we have isomorphisms $Ext_{R}^{i}(C, A) \cong Ext_{R}^{i}(D,B) $ for all $i>0.$ I have the feeling that using the long exact sequence derived from the short ones the cases where $i>1$ are easy. But I am having a lot of troubles with the case $i=1$ because $Hom(C, I)$ and $Hom(P, B)$ are not necessarily 0.
Asked
Active
Viewed 75 times
1 Answers
1
For all $R$-modules $X$, for all $k\ge 1$, one has an isomorphism $$\text{Ext}^k(X,B)\xrightarrow{(1)} \text{Ext}^{k+1}(X,A), $$ because $\text{Ext}^k(X,I) = \text{Ext}^{k+1}(X,I)= 0$.
For all $R$-modules $Y$, for all $k\ge 1$, one has an isomorphism $$\text{Ext}^k(C,Y) \xrightarrow{(2)}\text{Ext}^{k+1}(D,Y),$$ because $ \text{Ext}^{k}(P,Y)= \text{Ext}^{k+1}(P,Y)=0$.
Putting these together, (for $k\ge 1$) one has an isomorphism $$\text{Ext}^k(D,B)\xrightarrow{(1)} \text{Ext}^{k+1}(D,A)\xleftarrow{(2)} \text{Ext}^k(C,A).$$
peter a g
- 4,923
-
Thank you, so much! At some point I got lost and I knew it had to be a not too complicate one. Thanks again! – Ojotsk May 30 '19 at 22:47
-
My pleasure... But! The frustrating parts of this answer for me were: 1) there does not seem to be a native-TeX symbol to indicate that a map is an iso - I still can't believe it! One can build one - \xrightarrow{\sim} gives $\xrightarrow{\sim}$, but really, not so nice. 2) And then I wanted to put (1) and (2) on top (as I did in the answer). As you see, I gave up, used 'to', and said 'isomorphism' explicitly. [I wanted to indicate the 'natural' direction of the iso. ] – peter a g May 30 '19 at 23:52
-
Oops - I lied: not that anyone cares, but I did not use 'to.' I used \xrightarrow{(1)}, etc.... – peter a g May 31 '19 at 00:26