Approximation to half-iterate by high indexed natural iterates (base on ShlThrb) Gottfried Ultimate Fellow Posts: 900 Threads: 130 Joined: Aug 2007 08/21/2019, 09:48 AM (This post was last modified: 08/21/2019, 07:38 PM by Gottfried.) *Update, perhaps the key for the general solution, see marked paragrap below*             A funny aftermath to my previous thread. Let base $b$ be on the Shell-Thron-boundary, such that with $\small b=\exp(u\cdot\exp(-u))$ we choose some real $\small c$ such that $\small u=\exp(2 \cdot \pi \cdot I/c)$ .                                 I used $\small c=$ golden ratio $\small\approx1.61$                           We observed in the previous thread that with $\small z_0=a+0 I$ and $\small 0.5mindist , next() ); mindist = d; minh = h; minw = w; print( [ minh, minw, mindist ] ); ) } This gave the following table: Code: h z_h as approx to z_05 distance difference to next h ------------------------------------------------------------------------- [ 1, 0.0994504 - 0.793112*I, 0.821600] + 2 [ 3, 0.886975 - 0.365889*I, 0.173966] + 8 [ 11, 0.869856 - 0.584506*I, 0.0461214] + 34 [ 45, 0.883019 - 0.529344*I, 0.0106015] + 144 [ 189, 0.880502 - 0.542220*I, 0.00251854] + 610 [ 799, 0.881128 - 0.539172*I, 0.000593661] + 2584 [ 3383, 0.880982 - 0.539891*I, 0.000140194] + 10946 [ 14329, 0.881017 - 0.539721*I, 0.0000330925] + 46368 [ 60697, 0.881009 - 0.539761*I, 0.00000781224] +196418 [257115, 0.881011 - 0.539752*I, 0.00000184421] The differences of the iteration-heights are actually from the convergents of the cont-frac of c (read first row), and in steps of 3 (read star-markers): Code:cvgts of cont-frac of c * * * * * * * * * [1 1 2 3 5 8 13 21 34 55 89 144 233 377 610 987 1597 2584 4181 6765 10946 17711 28657 46368 75025 121393 196418 317811 514229 832040] [0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610 987 1597 2584 4181 6765 10946 17711 28657 46368 75025 121393 196418 317811 514229] So we get the surprising approximation using \( h(m)= 1 + \sum_{k=0}^m p$3k+2$$          where $p$k$$ is the numerator of the k'th convergent of the continued fraction of $c$ and then                        $\lim_{m\to \infty} |\exp_b^{0.5}(z_0)-\exp_b^{h(m)}(z_0)|=0$                      *Update*  The key for the observation is likely that because of the circular behave of powers of $u$ we have a modulo-situation in the background and that evaluates to $h=0.5$                          $\hspace{96}\small\lim_{m\to \infty} h(m)\equiv0.5\hspace{48}\pmod{\frac1c}$ It's similar when I used $\small z_0=0.8$ and I'd like to know whether this can be generalized to other fractional heights ... For instance here is the approximation to the $h=1/4$ fractional height: Code:z_025=schrI(s0 * u^0.25 )   \\ z_025= 0.843362 - 0.211533*I {w=0.7;   mindist=9;minh=-1;oldminh=0;   for(h=1,1 000 000,       w=exp(bl*w);       d=abs(w-z_025);       if( d>mindist,next());       mindist=d;minh=h;minw=w;       print([minh,minw,mindist,minh-oldminh]);       oldminh=minh) } This gave the following table. I've not yet an idea how to adapt the beginning such that the *diff(i_cv)* harmonize meaningfully: Code:            h     approx to z_025         dist       diff(h)   i_cv     diff(i_cv)  ///  here i_cv= index of diff(h) in cvgts-numerators -------------------------------------------------------------------------- [     1, 0.0994504 - 0.793112*I, 0.944266,          1]  1         1 [     2, 0.187985 + 0.00909452*I, 0.691517,         1]  2          3 [     3, 0.886975 - 0.365889*I, 0.160399,           5]  5         1 [     8, 0.792226 - 0.114578*I, 0.109614,           8]  6          3   [    16, 0.860617 - 0.255881*I, 0.0475870,         34]  9           2 [    50, 0.846561 - 0.219068*I, 0.00818594,        89] 11         1 [   139, 0.840795 - 0.205664*I, 0.00640565,       144] 12          3 [   283, 0.844382 - 0.213908*I, 0.00258456,       610] 15           2 [   893, 0.843542 - 0.211950*I, 0.000454229,     1597] 17         1 [  2490, 0.843220 - 0.211204*I, 0.000358160,     2584] 18          3 [  5074, 0.843419 - 0.211665*I, 0.000143838,    10946] 21           2 [ 16020, 0.843372 - 0.211556*I, 0.0000253073,   28657] 23         1 [ 44677, 0.843354 - 0.211515*I, 0.0000199633,   46368] 24          3 [ 91045, 0.843365 - 0.211540*I, 0.00000801522, 196418] 27           2 [287463, 0.843363 - 0.211534*I, 0.00000141031, 514229] 29         1 [801692, 0.843362 - 0.211532*I, 0.00000111253,       ]         cvgts  1 2 3 4 5 6  7  8  9 10 11  12  13  14  15  16   17   18   19   20    21    22    23    24    25     26     27     28     29     30 index k  * *     * *        *     *   *           *        *    *               *           *     *                   *             *      *  marked [1 1 2 3 5 8 13 21 34 55 89 144 233 377 610 987 1597 2584 4181 6765 10946 17711 28657 46368 75025 121393 196418 317811 514229 832040] numerator [0 1 1 2 3 5  8 13 21 34 55  89 144 233 377 610  987 1597 2584 4181  6765 10946 17711 28657 46368  75025 121393 196418 317811 514229] denominator` Well, if this works more generally, then I'd conclude: the properties of the Siegel-disc with that type of exponential bases supports the meaningfulness of the Schröder-mechanism for the computation of the fractional iterates in that cases. (my time is reduced at the moment, perhaps I can come back to this later) Gottfried Helms, Kassel tommy1729 Ultimate Fellow Posts: 1,906 Threads: 409 Joined: Feb 2009 09/09/2019, 10:50 PM I am confused by your sum formula and use of continued fractions. I Will post a thread in a few min to explain how I do or see dynamics of a Siegel disk. Maybe you can clarity and compare. Regards  Tommy1729 « Next Oldest | Next Newest »

 Possibly Related Threads… Thread Author Replies Views Last Post [GEN DYN] Iterates of tanh(A x)^(2m+1) tommy1729 0 32 05/22/2023, 10:23 PM Last Post: tommy1729 Has anyone studied the iterates of $\frac{1}{e^{-1/z}}$ Daniel 2 59 05/22/2023, 08:56 PM Last Post: tommy1729 Simple limit approximation to exp(x) tommy1729 0 29 05/16/2023, 11:13 PM Last Post: tommy1729 [2sinh] exp(x) - exp( - (e-1) x), Low Base Constant (LBC) 1.5056377.. tommy1729 3 131 04/30/2023, 01:22 AM Last Post: tommy1729 Discussing fractional iterates of $f(z) = e^z-1$ JmsNxn 2 480 11/22/2022, 03:52 AM Last Post: JmsNxn Half-iterate exp(z)-1: hypothese on growth of coefficients Gottfried 48 6,966 09/09/2022, 12:24 AM Last Post: tommy1729 Base -1 marraco 15 20,415 07/06/2022, 09:37 AM Last Post: Catullus Imaginary iterates of exponentiation jaydfox 9 14,215 07/01/2022, 09:09 PM Last Post: JmsNxn Natural Properties of the Tetra-Euler Number Catullus 6 1,234 07/01/2022, 08:16 AM Last Post: Catullus I thought I'd take a crack at base = 1/2 JmsNxn 9 3,122 06/20/2022, 08:28 AM Last Post: Catullus

Users browsing this thread: 1 Guest(s)