22 March 1999: New c-coded files are now available 24 March 1999: Fixed sin scaling in spinv.c 25 March 1999: Fixed underflow problem in ordblk (ulib.c), fixed potential problem with reading option lines in calfit, fixed DIAG=2 bug in spinv. 22 April 1999: Fixed bug in Itot option in spinv.c when more than 2 spins. 25 May 1999: Fixed bug in calfit involving blends when there are bad quanta in .lin files; fixed bug in spinv that appears when largest number of spins is not the default; added feature to calculate partition funtion in calcat for temperature specified in line 2 of .int file. 23 July 1999: Fixed bug in spinv.c that caused IQNFMT to have wrong 10's digit when F was half integer 7 Sept. 1999: Fixed bug in spinv.c that caused crash for large spin sets 13 Sept. 2000: Many small revisions including: (1) negate Hamiltonian for kappa > 0 on first pass to improve logic for oblate case, (2) make separate ORDHAM and FIXHAM, (3) subtract constant from energy for more precision, (4) special code for unit matrix, (5) fix up phase for cross-multiplet interaction, (6) efficiency improvement for 3J, 6J, and 9J routines, (7) eliminate use of .unf file, (8) size check on correlation matrix elements in .par and .var files. 9 Oct. 2000: Changed definiton of dipole type = 10 (see spinv.html) 13 June 2001: Include sort of .cat file at the end of spcat and dpcat 7 July 2001: Fix bug for multiple l doubled states 26 Sept. 2001: Fix bug for oblate interactions of type 6x and 2x Split readqn and pcard from ulib and improve code. 10 Dec. 2003: Augmentation for n-fold symmetric tops with n = 2,3,4,5,6 Improved portability with changes proposed by splint. 10 Mar. 2004: New code for resolved hyperfien for n-fold symmetric tops using Itot basis for a single set of n-equivalent spins. 24 May 2004: Fix bugs for v > 99, use BCD internally for parameter and dipole identifiers 20 July 2004: internal BCD modification had bugs for SPFIT for dipoles with SI1 or TYP fields in IDIP not zero. 21 July 2004: fixed bug in SPFIT/SPCAT that affected default VSYM 24 Aug. 2004: format of frequency field in .cat field is now expandable in spcat, dpcat, and calmrg to accomodate IR frequencies in MHz 30 Dec. 2004: expanded Euler operators 29 May 2005: fixed bug in operators connecting l=0 to l-doubled states. fixed minor memory leak for DIAG=1 cases. 31 May 2005: fixed bug so idpar field in par file can be followed by decimal point to retain compatibility with older files 8 July 2005: dpfit had bug when speciying parameter 27, code and doc fixed 9 July 2005: changed centrifugal distortion parameters for hyperfine and quadrupole. Earlier versions had bug that suppressed applying centrifugal distortion to Delta J ne 0 matrix elements. 28 Sept. 2005: fixed a bug in SPFIT/SPCAT for Itot basis for matrix elements with Delta Itot odd. The signs for some off-diagonal elements have the wrong sign. 27 Oct. 2005: fixed minor bugs in magnetic dipole intensities. fixed bug in dipole TYP = 8 and 12 that produced dipole with wrong symmetry except for l-doubled states. 23 Dec. 2005: fixed bugs in Itot basis for l-doubled states. several poorly debugged versions were put on the web prematurely in November and December. 3 Jan. 2006: new expanded options for fitting blended lines that allows fitting the rms frequency separation as well as the average frequency. See spinv.pdf documentation. 16 Jan. 2006: fixed bug in intensity calculation that resulted in improper check of statistical weights. fixed bug in quantum format field 26 Jan. 2006: fixed sign error for dipoles 11 Feb. 2006: Fixed bug that made some derivatives a factor of 2 too large. Impact on fitting and estimated uncertainties. 9 May 2006: Extended definition of EWT1 for l-doubling situations 11 May 2006: Fixed bug concerning nvib > 99 and in special cases nvib > 9. 26 May 2006: Fixed least square fit for rank-decicient problem, i.e. specify the same parameter twice. Fixed case of I = 0, Itot = 0. 22 June 2006: Fixed bugs in calmrg that produced incorrect .bad file 15 July 2006: Fixed bug in diagonalizer when of row of off-diagonal elements was zero 18 July 2006: Fixed bug in spfit that first parameter had wrong step for fit 17 Aug. 2006: Improve checking of spin weights in spinv 16 Oct. 2006: Increase options for phases between Wang states. New option for specifing Lz operator explicitly. 22 Mar. 2007: Improvements for handling internal rotation. Update documentation. 1 June 2007: Gup MSB can be a character in cat and mrg files (gup is < 3600) 4 June 2007: Fix bug in calmrg when merging two unequal lines 9 June 2007: Fix bug in SPCAT with IDIAG = 0, no egenvector transformation of S 11 July 2007: Fix bug that rejected Euler denominator parameters 12 July 2007: Fix bug that rejected parameters for Itot basis 3-fold sym rotors 19 July 2007: Enhancement of DPFIT and SPFIT .par file in line 2 FACT field, a negative value makes parameter errors follow convention. 24 July 2007: Improved diagnostic messages when parameter or dipole is rejected 6 Nov. 2007: Fixed bug that did not include Itot = 0 lines in .cat files and .str files 6 NOv. 2007: Fixed bug in ORDBLK(in ulib.c) that caused bad quanta for coupled N = 0 states