Open mathlusiverse opened 3 months ago
It seems that maxima returns a factorisation instead of all solutions:
sage: ex=2048*x^10 + 4096*x^9 + 4096*x^8 + 2560*x^7 + 1280*x^6 + 544*x^5 + 192*x^4 + 48*x^3 + 12*x^2 + 2*x + 1 == 0
sage: solve(ex,x)
[0 == 128*x^5 + 128*x^4 + 64*x^3 + 16*x^2 + 8*x + I*sqrt(7) + 1, 0 == 128*x^5 + 128*x^4 + 64*x^3 + 16*x^2 + 8*x - I*sqrt(7) + 1]
and then our conversion fails on that maxima output.
Steps To Reproduce
The following will cause exception in Sagemath:
But if I change the coefficient of x^10 from 2048 to 2047 or 2049, Sagemath can solve it.
The original equation can be solved using the public account of Mathematica Alpha, which gives the following approximate solutions (they are conjugate pairs, only five are shown here):
Expected Behavior
Expect to see 5 pairs of conjugate complex numbers.
Actual Behavior
It throws the following exception:
Additional Information
I changed many coefficients slightly, Sagemath can solve them. Sagemath can handle similar 10th degree polynomial equations. The roots of the original and slightly modified 10th degree equations have similar (visual) behavior, they spread out quite nicely on the complex plane.
Environment
Checklist