Implementation of cTPQ (canonical TPQ) method

[tmisawa]

1. I implemented the cTPQ method as follows:
   • |Φ(β=N*Δτ)> = U(Δτ)^N |Φ_initial>
   • U(Δτ) = exp[-Δτ/2 H] ~ sum_{k=0}^{k_max} (-Δτ/2*H)^k/k! (k_max-th Taylor expansion of exp[-Δτ/2 H])
   • k_max = ExpandCoeff, Δτ=1.0/LargeValue

In caclmod.def, by specifying,

CalcType   5

one can use cTPQ method.

2. I also added an option for choosing the initial vectors of the TPQ calculations (mTPQ and cTPQ):

In caclmod.def, by specifying,

InitialVecType   -1

one can generate the uniformly distributed random vectors on the 2N_H dimensional unit supersphere (N_H is the dimension of the Hilbert space).

Up to now, as the initial vector, we employ the vectors whose components are given by the uniform random complex numbers.

[tmisawa]

@yomichi マニュアルも更新しました。