Successive modifications of limit periodic continued fractions (Q1094805): Difference between revisions

From MaRDI portal
RedirectionBot (talk | contribs)
Removed claim: reviewed by (P1447): Item:Q588676
RedirectionBot (talk | contribs)
Changed an Item
Property / reviewed by
 
Property / reviewed by: Peter Wynn / rank
 
Normal rank

Revision as of 04:41, 20 February 2024

scientific article
Language Label Description Also known as
English
Successive modifications of limit periodic continued fractions
scientific article

    Statements

    Successive modifications of limit periodic continued fractions (English)
    0 references
    0 references
    1987
    0 references
    b\({}_ 0\) and \(a_ i\) (i\(\geq 1)\) being complex numbers, let C be the value of the continued fraction (*) \(b_ 0+\frac{a_ 1}{1+}\frac{a_ 2}{1+}..\). with convergents \(C_ n=A_ n/B_ n\). \(S_ n(r_ n)=(A_ n+r_ nA_{n-1})/(B_ n+r_ nB_{n-1}),\) where the converging factor \(r_ n\) is an approximation to the value of the tail \(\frac{a_{n+1}}{1+}\frac{a_{n+2}}{1+}..\). of (*), is an approximation to C. The case in which lim \(a_ n=a\not\in [-\infty,-1/4]\) is considered in detail. \(r_ n\) is taken to be the root of smallest modulus of the equation \(x^ 2+x-a=0\). The approximations \(S_ n(r_ n)\) (n\(\geq 1)\) are exhibited as convergents of a continued fraction which in turn is treated in the same way. Sequence to sequence transformations \(\{C_ n\}\to \{D_ n\}\) of the form \(D_ n=(g_ nC_{n-1}-g_{n- 1}C_ n)/(g_ n-g_{n-1})\) are dealt with in terms of the above theory: the \(C_ n\) are taken to be convergents of (*) and then \(D_ n=S_ n(r_ n)\), where \(r_ n=-B_ ng_ n/B_{n-1}g_{n-1}.\) Aitken's \(\delta^ 2\)-process, for which \(g_ n=C_{n+1}-C_ n\), is studied in detail.
    0 references
    limit periodic continued fractions
    0 references
    Aitken's delta square algorithm
    0 references
    convergence acceleration
    0 references
    0 references

    Identifiers