Abstract:Newton's identities provide a way to express elementary symmetric polynomials in terms of power polynomials over fields of characteristic zero. In this article, we study the failure of this relation in positive characteristic and what can be recovered. In particular, we show how one can write the elementary symmetric polynomials as rational functions in the power polynomials over any commutative unital ring.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: in positive characteristic, the failure of Newton's identities to express the relationship between elementary symmetric polynomials and power polynomials, and to explore which relationships can be restored in this case. Specifically, the paper discusses how to represent elementary symmetric polynomials as rational functions of power polynomials in any commutative unital ring.
### Detailed Explanation:
1. **Background Introduction**:
- In a field of characteristic zero, Newton's identities provide a method to represent elementary symmetric polynomials with power polynomials.
- However, in the case of positive characteristic, this relationship fails. For example, in a field of characteristic \( r > 0 \), the relationship between power polynomials \( p_k \) and \( p_{kr} \) leads to the non - establishment of certain algebraic relationships.
2. **Main Problems**:
- **Problem 1**: In a field of positive characteristic, can all power polynomials generate the ring of symmetric polynomials?
- The paper shows by an example that the answer is no. For example, in the two - variable symmetric polynomials over \( F_2 \), \( e_2 \) cannot be generated by power polynomials.
- **Problem 2**: In a field of positive characteristic, can elementary symmetric polynomials always be represented as rational functions of power polynomials?
- Schönhage proved that the answer to this question is yes, that is, for any field \( K \), we have \( K(e_1,\ldots,e_n)=K(p_1,p_2,\ldots) \).
3. **Main Contributions**:
- The paper gives a new proof, which is applicable to any base ring \( R \), that is:
\[
Q(R[e_1,\ldots,e_n]) = Q(R[p_1,p_2,\ldots])
\]
where \( Q(-) \) represents the total fraction ring.
- The paper also studies the sub - algebra generated by power polynomials and proves that in the case of positive characteristic, this sub - algebra is not finitely generated.
4. **Algorithms and Applications**:
- It provides a simple algorithm for representing elementary symmetric polynomials as rational functions of power polynomials.
- It gives an explicit description of the denominator, which is helpful for determining whether the value of an elementary symmetric polynomial can be calculated under given conditions.
5. **Numerical Valuation**:
- It explores how to determine the coefficients of the characteristic polynomial according to the valuation of power polynomials in a discrete valuation ring.
### Summary:
The core problem of the paper is to solve the failure problem of Newton's identities in positive characteristic fields, and through new proofs and algorithms, it shows how to represent elementary symmetric polynomials as rational functions of power polynomials in any commutative unital ring. This not only extends the scope of application of Newton's identities but also provides new insights into the theory of symmetric polynomials in positive characteristic fields.