A fast numerical algorithm for the composition of power series with complex coefficients (Q1095657): Difference between revisions

From MaRDI portal
Set OpenAlex properties.
ReferenceBot (talk | contribs)
Changed an Item
 
Property / cites work
 
Property / cites work: \(0(n^{2.7799})\) complexity for \(n\times n\) approximate matrix multiplication / rank
 
Normal rank
Property / cites work
 
Property / cites work: The complexity of partial derivatives / rank
 
Normal rank
Property / cites work
 
Property / cites work: Fast Algorithms for Manipulating Formal Power Series / rank
 
Normal rank
Property / cites work
 
Property / cites work: Fast modular transforms / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3325040 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Fast multiplication of large numbers / rank
 
Normal rank
Property / cites work
 
Property / cites work: An algorithm for division of powerseries / rank
 
Normal rank
Property / cites work
 
Property / cites work: Die Berechnungskomplexität von elementarsymmetrischen Funktionen und von Interpolationskoeffizienten / rank
 
Normal rank

Latest revision as of 12:38, 18 June 2024

scientific article
Language Label Description Also known as
English
A fast numerical algorithm for the composition of power series with complex coefficients
scientific article

    Statements

    A fast numerical algorithm for the composition of power series with complex coefficients (English)
    0 references
    0 references
    1986
    0 references
    Let P, Q be formal power series in a variable t over an infinite field K, \(P(0)=0\). The composition problem is the task of computing the coefficients of the polynomial \(Q\circ P\) mod \(t^{n+1}\) given the first \(n+1\) coefficients of P and Q. The best known upper bound for the Ostrowski complexity of this problem is reached by an \(O(n^{3/2}(\log n)^{1/2})\) algorithm due to Brent and Kung. On the other hand no nonlinear lower bound is known so far. The author shows that the composition problem is essentially equivalent to computing only the nth coefficient of \(Q\circ P\). A second result determines the complexity of an approximate version. It is shown that the complexity of computing \(Q\circ P\) mod\(\prod^{n}_{i=0}(t-\epsilon_ i)\) \((\epsilon_ 0,...,\epsilon_ n\in K^{\times}\) pairwise different) is of order n log n.
    0 references
    formal power series
    0 references
    composition problem
    0 references

    Identifiers