Consider the set $\Bbb{K}[[x]]$ this MSE question indicates that this is the set of expressions of the form: $$\sum_{n=0}^\infty a_nx^n\quad \text{where} \quad a_n \in \Bbb{K} $$ I don't understand the meaning of the terms $x^n$. In the most general case what space/field do they live in? and what does the exponent $n$ mean?
-
Do you know about formal polynomials? if you do, it's similar, with the difference that we allow the $a_n$ to be non-zero for arbitrary values of $n$. – Xam Aug 01 '17 at 19:10
-
First figure out what $x$ is in $K[x]$. – GEdgar Aug 01 '17 at 19:12
-
@Xam Unfortunately not, I came across the concept of 'formal' things today when looking for a definition of the Laurent series. I am finding it really hard to find any information on these things (at least in the general case). – Quantum spaghettification Aug 01 '17 at 19:13
-
1@Quantumspaghettification this: https://en.wikipedia.org/wiki/Polynomial_ring and this: https://en.wikipedia.org/wiki/Formal_power_series might help you. – Xam Aug 01 '17 at 19:17
1 Answers
Let $R$ be any ring. Then $R[[X]]$ denotes the ring of formal power series over $R$.
Informally, one often simply defines it as the set $$\Big\{\sum_{n=0}^∞a_nX^n;~a_n ∈ R~\text{for all}~n∈ℕ_0\Big\},$$ and then proceeds to say “Its elements are formal expressions” (without clarifying what formal expressions are) and “This yields a ring when equipped with the suggestive addition and multiplication on it.” One then calls “$X$” a formal variable and leaves it undefined and uninterpreted.
To the formalist, this is of course extremely unsatisfactory. But fortunately, the wikipedia article above gives a formal definition of it here.
Following this definition, “$\sum_{n=0}^∞ a_nX^n$” simply becomes a suggestive notation for the sequence $(a_n)_{n ∈ ℕ_0}$. Again, this might be unsatisfactory. However, if you define $X$ to be the special sequence $$X = (0,1,0,…),$$ this notation actually makes sense: One may interpret $R[[X]]$ as topological ring which is complete in a certain sense and then the formal series literally becomes $$\sum_{n=0}^∞a_nX^n = \lim_{N→∞} \sum_{n=0}^N a_nX^n.$$ If you are interested in this viewpoint, I suggest you first read wikipedial on topological rings, specifically the examples section, discussing the $I$-adic topology. You may read further about this in any book on commutative algebra such as Atiyah–Macdonald’s book.
By the way, this trick also works for polynomial rings $R[X]$ (with which I presume you are at least somewhat/intuitively familiar), defining $R[X]$ as a set of sequences in $R$ and $X = (0,1,0,…)$: You just have to restrict yourself to sequences that are nonzero only at finitely many terms. But then, the polynomial $\sum_{n=0}^N a_nX^n$ too really is the sum of monomials $a_nX^n$ and $X^n$ really is the product of some special element $X = (0,1,0,…)$ by itself.
- 18,539
-
Are you saying that $X \in R[X]$? If so how is the product $a_1 X$ defined such that both $a_1 X \in R[X]$ and $X \in R[X]$? (since we may be considering an arbitrary field) – Quantum spaghettification Aug 02 '17 at 06:12
-
1@Quantumspaghettification Yes, $X ∈ R[X]$. For any $a ∈ R$, we do also simply write “$a$” for the sequence $(a,0,0,…)$. By the way multiplication is defined on $R[X]$, this makes $$aX = (a,0,0,…)(0,1,0,…)= (0,a,0,…) = a·X$$ if you interpret $a·X$ to be the classical scalar multiplication on sequences. As you can see, the product $aX$ is defined in no other way than the general product of two polynomials (or power series). – k.stm Aug 02 '17 at 06:16