MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $A$ be a commutative ring (not necessarily noetherian).

Let $I\subseteq J\subseteq A\,$ be two finitely generated ideals.

Let us denote the completion functor by $\Lambda_K (M) = \varprojlim_n M/K^nM$.

I would like to compare the two rings:

$\Lambda_J(A)$ and $\Lambda_{J'} (\Lambda_I(A))$ where $J' = J\Lambda_I(A)$.

Are they the same? are they isomorphic?

I thought about showing that in the right hand side, $A$ is dense in the $J$-adic topology, and that it is $J$-adically complete. Is this true? and are those two facts enough to show that it is isomorphic to the left hand side?

Edit: I should mention that in my application, the two rings $\Lambda_I(A)$ and $\Lambda_J(A)$ are noetherian.

share|cite|improve this question

Let $A$ be a commutative ring, $I\subset A$ be an ideal. Denote by $\hat{A}$ the completion with respect to the $I$-adic filtration.

One can consider the kernel $\hat{I}_n$ of the natural projection $\hat{A}\to A/I^n$. The obvious observation is: $\hat{A}/\hat{I}_n=A/I^n$. Thus, $\hat{A}=\varprojlim_n\hat{A}/\hat{I}_n$.

Now, it' clear that $I^n\hat{A}\subseteq\hat{I}_n\subseteq(\hat{I}_1)^n$. In Noetherian case they are always equal. This gives an affirmative answer to your question. In general they can differ. However, I can not come up with an example immediately.

share|cite|improve this answer
Just to be sure, are you claiming that my question has a negative answer in general? – DARK Oct 10 '12 at 9:49
If $I$ is of finite type then $\widehat{I} = I \widehat{A}$ (more generally if $M$ is a f.g. $A$-module then $M \otimes \widehat{A} \to \widehat{M}$ is surjective. This is indeed false in general, for example you can take $A=k[X_1,X_2,\ldots]$ and $I=\langle X_1,X_2,\ldots \rangle$. Then $F = X_1+X_2^2+\cdots \in \widehat{I}$ doesn't belong to the image of $I$. – François Brunault Oct 10 '12 at 11:09

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.