CCL:G: Tozer's diagnostic test in Gaussian

 Sent to CCL by: "J.  Phoenix" []
 Dear CCLs,
 I am trying to compute the Tozer's lambda parameter that relates the spatial
 overlap between orbitals in a certain excitation using TDDFT calculations (JCP,
 128, 044118 (2008)). The procedure I am following is to extract the overlap
 matrix of AOs in the Gaussian output and then compute for each MO pair:
 <Psi_1|Psi_2> = sum_i sum_j c_1_i c_2_j <X_i|X_j>, using the MO
 coefficients. When coefficients are taken with their sign, I get an overlap of
 one between the same MO, and zero between different MOs. However, when I use the
 absolute value of the coefficients to compute the inner product of the moduli,
 overlaps larger than 1 are obtained in some cases.
 Is there anything wrong in my procedure for computing the inner product of the
 moduli of two orbitals or is not expected to be a value between 0 and 1?
 Thanks in advance,