OFFSET
0,3
COMMENTS
Coefficients of the numerator polynomials of the bivariate Maclaurin series ("inverse Kepler equation") developed as Lagrange inversion E = KeplerInv(e,M) of Kepler's equation M = Kepler(e,E) = E - e*sin(E), where e=numeric eccentricity, M=mean anomaly, E=eccentric anomaly. The series is KeplerInv(e,M) = M/(1-e) + Sum_{n>=1} (-1)^n*(Sum_{j=1..n} a(n,j)*e^j)/(1-e)^(3n+1)*M^(2n+1)/(2n+1)! = M/(1-e) - (e/(1-e)^4)*M^3/3! + ((e+9*e^2)/(1-e)^7)*M^5/5! - + ... .
The element a(n,n) with highest index in each row (the diagonal element) has the form Product_{j=1..n} (2*j+1)^2.
The derivative dKepler/dE = 1 - e*cos(E) goes to zero at E = i*arccosh(1/e) in the complex plane. Thus dKeplerInv/dM goes to infinity at M = i*(arccosh(1/e) - sqrt(1-e^2)), so that the radius of convergence of KeplerInv(e,M) is arccosh(1/e) - sqrt(1-e^2). KeplerInv(e,M) converges linearly within the circle of convergence |M| < arccosh(1/e) - sqrt(1-e^2).
LINKS
Wikipedia, Kepler's equation
FORMULA
EXAMPLE
Matrix (regular triangle) lexicographically ascending in the rows:
1;
1, 9;
1, 54, 225;
1, 243, 4131, 11025;
1, 1008, 50166, 457200, 893025;
1, 4077, 520218, 11708154, 70301925, 108056025;
1, 16362, 5020623, 243313164, 3274844175, 14427513450, 18261468225;
...
CROSSREFS
KEYWORD
nonn,tabl
AUTHOR
Herbert Eberle, Feb 23 2019
STATUS
approved