Skip to content

tk-yoshimura/MathieuFunction

Repository files navigation

MathieuFunction

Definition

Eigenvalues

eigen summary

Continued Fraction and Convergence

DLMF 28.6

a even
b even
a odd
b odd

Normalized

eigen normalized

Average and Difference

eigen mean sub
eigen mean raw
eigen sub raw

Rational Coefficient Series

eigen rcoef md
eigen rcoef md grad0
eigen rcoef m
eigen rcoef d

Numeric Table (64 digits)

Continued Fraction Terms

128bit
eigen cfrac 128bit
256bit
eigen cfrac 256bit
512bit
eigen cfrac 512bit
1024bit
eigen cfrac 1024bit

Fourier Series

Mathieu function with eigenvalue a for q is a periodic solution and can be expressed as a Fourier series.
coef fourier

Coefficient Matrix

The relationship between the coefficients is expressed using a tridiagonal matrix T.
coef matrix
coef param

Backward Recurse

Backward recursion is used as an approximate method to solve the coefficient sequence.
coef backrecur

Solve Matrix

In backward recursion, terms may cancel each other and produce a value close to zero, although this is a rare case.
In this case, it must be solved in pieces as a linear problem because the digits drop out and subsequent values become inaccurate.
coef det
coef solve

When q is small, the accuracy of calculation is extremely poor for ce, se(n=1), se(n=2), where the a - r0 is close to zero.
Therefore, it is necessary to separately prepare an approximation of zero-shifted eigenvalue function.

Implement

DoubleDouble (30 digits)