search

alexjbest / cluster-pictures

This package implements the machinery of cluster pictures of Maistret, Morgan, Dokchitser and Dokchitser in Sage.
https://alexjbest.github.io/cluster-pictures/
GNU General Public License v2.0
3 stars 4 forks source link

sigmaK(P) != P when we expect it to be #26

Closed rbommel closed 4 years ago

rbommel commented 4 years ago

Happens for the Tamagawa number for curve 588.a.18816.1.

So I believe what happens, is the following. The cluster picture is (****(**)) and Frobenius acts cyclically on the first four roots, and cyclically on the two in the twin. Now the subcluster (**) is invariant under Frobenius, but its theta_squared is not, because the center [sic] of (**) is not invariant under Frobenius.

rbommel commented 4 years ago

Appeared to be a problem with the Teichmuller lift. The number of terms there was not correct.