updates-3.3

04/06/07: Update documentation. -CDS and ALR

04/03/07: Added SCS-MP2 to the "mp2" module.  -CDS and TT

12/03/06: Changed configure to be able to compile Psi3 without a fortran compiler.  -EFV

08/02/06: Added dipole moment and transition dipole moment analysis 
within DETCI.  Unified CI and CC code to analyze multiple density
matrices in OEPROP. -CDS

07/28/06: Given the improvements to libpsio to avoid file corruption, I
added restart options to ccresponse.  The code can now recover almost
exactly where it left off (to within an iteration of each perturbed wave
function for a given response tensor) prior to a crash.  This should save
vast amount of CPU time for large optical rotation and polarizability
calcualtions. -TDC

07/27/06: Significant changes to libpsio designed to avoid corruption of
files in cases of program crashes due to external problems (e.g. power
failures).  The old file structure placed the entire table-of-contents (TOC)
at the end of the file (usually only when the file was closed).  The new
structure stores the TOC data for each entry/record as a header to the data
itself.  This header is updated if necessary just before a given write.
Thus, if the program crashes, the TOC data should still be accessible and
recovery of the calculation may be possible. -TDC

07/25/06: Update Parameters.S in DETCI to only use time-reversal
symmetries when S is an integer. -CDS

07/21/06: Added transqt2 to the repository.  This will soon be a full
replacement for transqt with substantially reduced disk-space requirements
and better speed.  -TDC

06/26/06: Multithreaded RCI DBOC code.  -EFV

06/21/06: DBOC now uses MEMORY keyword and can do RCI DBOC calculations where
string overlaps are kept on disk.  -EFV

06/14/06: Modified DETCI, DETCAS, and DETCASMAN to prepend prefix to names of 
all intermediate files.  -EFV

06/12/06: Modified DBOC code to avoid computation of the wave function at the
reference (non-displaced) geometry. This should simplify computations in
high symmetries.  -EFV

11/29/05: Added ability to carry out Sekino-Bartlett Model III calculations
of frequency-dependent linear response.  This is a simplified EOM-like
approach that is size extensive.  See Adv. Quantum Chem. 35, 149 (1999). -TDC

11/28/05: Updated configure.in for autoconf-2.57.

11/15/05: Implemented simultaneous length/velocity gauge optical rotation
calculations for CC2 and CCSD.  Also extended the perl testing scripts to
check these types of calcs as well as multi-wavelength calcs.  -TDC

11/12/05: Implemented code to compute and write to disk symmetry-adapted
first-order derivative integrals with respect to Cartesian nuclear 
displacements.  Symmetry-adapted linear combinations are computed by input 
and stored in checkpoint file. -EFV

09/10/05: Major improvements to <0|L*(HBAR*X1*Y1)c|0> terms of the linear
response function in ccresponse.  Most contractions are now taken from
results from cclambda, and other terms are computed independent of the
perturbations.  The final contractions are just simple dot products.  This
completely eliminates the need for <ab|cd> terms in the response function. -TDC

08-09/05: Added streamlined <ab|cd> algorithm to all CC codes, including
ccenergy, cclambda, cceom, and ccresponse.  The new algorithm appears to be
significantly faster than the previous one, which is not surprising because
its prefactor is 1/4 of the old one. -TDC

06-07/05: Removed all iabc-based bottlenecks from CCSD codes including
EOM-CCSD and CCSD-LR.  The only remaining bottlenecks occur in CCSD
gradients (ccdensity).  Also much reduction in disk-space requirements. A
new test case (cc44), tests most (all?) of the out-of-core algorithms with
an aug-cc-pVTZ/RHF-EOM-CCSD calculation on the lowest A1 and B2 excited
states of H2O using only 2.0 MB of memory.  -TDC

05/25/05: Added RASSCF and SA-CASSCF capabilities.  Replaced wfn=DETCAS
with wfn=CASSCF or wfn=RASSCF.  Added preliminary implementation of
frozen vs restricted orbitals in MCSCF (works but can only do either/or
for now).  Automatically set proper ex_lvl if fci=true, automatically
set fci=true if wfn=CASSCF.  -CDS

01/03/05: Added capability to compute and print pair correlation
energies at MP2 and CC levels (RHF cases only).  -EFV

09/17/04: Added frequencies by finite differences of energies and gradients
(3- and 5-point formula) using symmetry-adapted cartesian coordinates.
Added chkpt_override_occ() to libchkpt to keep DOCC from getting parsed.
Added jobtype=FC and SYMM_FC to use internal coordinates if desired.
Added capability to do numerical, constrained optimizations.
Changed many of the test cases.  -RAK

08/12/04: Correct <S^2> computation for UHF wfns.  -TDC

07/09/04: Added UHF-based CC3 energy code, plus a test case (cc33).  This
code uses DGEMM calls, but is slow in part because it fails to limit IJK
combinations in the triples code to unique T3's.  I'll add this later, but
for now, the current version works.  -TDC

06/10/04: Added CC3 energy code to ccenergy plus a test case (cc32).  This
code stills needs conversion to a DGEMM-based version for greater
efficiency.  -TDC

06/08/04: Adding CCSD optical rotation test cases to main trunk, along with
new code for velocity-gauge dipole-based optical rotation for
origin-independent results. -TDC

05/03/04: Merged code from psi-3-2-1 into main trunk. Fixed a number of
small bugs arising from the merge.  -TDC

4/26/04: Standardized naming of cchbar matrix elements Wmnie and Wamef.  Spin-
adapted computation of the Xi amplitudes in ccdensity for RHF references.
Added test cases for all types of EOM gradients. -RAK

3/19/04: Finished bulk of spin-adaptation of cchbar codes, including the
Wmbij and Wabei terms, which are by far the worst.  This version should use
considerably less disk space and be 2-3 times faster than the psi-3.2
version of cchbar.  -TDC

3/17/04: Substantial modifications to cchbar, cclambda, cceom, and 
ccresponse to take advantage of RHF spin-adaptation.  Also eliminated
several redundant storage arrangements of Wamef elements. Added an
ROHF-EOM-CCSD test case, cc22.  -TDC

3/17/04: Modified DBOC code to take into account symmetry of the molecule
as much as possible to reduce the number of wave function computations. -EFV

3/16/04: Finished modifications to DBOC to allow 4-point formula for derivative
wave functions. 4-point results seem much more accurate and independent of
the displacement size. -EFV

3/15/04: Started reorganization of DBOC code to allow for 4-point gradient formula.
Currently the code works again for 2-point displacements. -EFV

3/10/04: Corrections: (1) fixed ccresponse for new storage of lambda
amplitudes; (2) added cc21 test case to tests/Makefile.in and configure.in.
-TDC

3/2/04: Added new EOM-CCSD gradient code, including test case cc21. -RAK

1/26/04: Fixed a bug in src/lib/libdpd/contract444.c for out-of-core cases.
I was not including the size of a single file4 row on the X component of
the X*Y-->Z contraction, which is necessary for some <ab|cd> cases in
cclambda.  This fix also appears on the psi-3-2 branch. -TDC

1/16/04: Added documentation for new MP2 features to the user's manual
and the man pages. -MLA

1/14/04: Added UHF and ROHF MP2 one-particle density matrices. -MLA

1/13/04: Added code 'giao' to compute B-field derivative Fock matrices.  -EFV

1/13/04: Checked in initial implementation of derivative integrals over GIAO Gaussians
necessary for optical rotation and NMR chemical shift (minus certain derivative
one-electron integrals) calculations.  -EFV

1/13/04: Added semicanonical orbitals to transqt. The reference changes from 
ROHF to UHF so I changed all the codes that "should" be affected.  Also added
UHF and ROHF MP2 energy including frozen orbitals. -MLA -TDC