Attempt to make own implementation of "Kneser" algorithm: trouble

[mike3]

how do I change base (from exp(1))??

See where it says "base" in the Initialize() part? This needs to be changed. It should work for real bases greater than \( \eta = e^{1/e} \). Not sure how big you can make the base due to the fixed initial fixpoint approximation, though.

ok. I edited base from exp(1) to 10 in your code. I entered 'Initialize()', 'DoKneserIteration(12)', 'FullTetApprox(0.5)' and can evaluate result. but... if base<eta then it doesn't work, ok?
but after I tested your code, Kneser.gp by Sheldon is much better precision than your code.

Yes, that's right. But if I can improve it with the just-suggested algorithms, I should be able to get it to go to full precision...