Dear Tenpy developers

I was playing around with purification method for Finite temperature on a 2-d lattice Kitaev Honeycomb model.

I tried following checks for my model.

1.) I checked the energy of model at infinite temperature which comes out to be zero, which is correct also(as trace of hamiltonian is zero).

2.) But at small \beta ~ 10^{-3} or even smaller \beta, energy of my model does not match with ED result.

3.) If I change the coupling one particular bond (\sigma_z*sigma_z) lying on z bond of honeycomb lattice to any value, I see no change in <E(beta=10^{-3})> (energy of model).

It looks to me there is a bug in PurificationApplyMPO.

Please let me know.

Thanks