-
Notifications
You must be signed in to change notification settings - Fork 3
LBNL-AMO-MCTDHF v1.36! . . . . .
License
LBNL-AMO-MCTDHF/V1
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
** LBNL-AMO-MCTDHF VERSION 1.36 ** v1.36 soft coulomb potential ironed out 1d version v1.35 sech^2 potential ironed out 1d version v1.34 contact operator available polyatomic version. Set use_contactop=.true. in namelist &sincparinp. Hydrogenic eigenvalues 1s, 2s, 2p should closely match actual values nuccharge^2/2 and nuccharge^2/8 up to nuccharge * spacing = 3 . Actual improvement to simulation results is not yet tested. v1.34 timing files report in units of seconds v1.34 much less memory waste for restricted configuration spaces (still not perfect, about 3x memory waste not 20x for some cases, still 3x waste of time.. now time and memory are wasted proportionally) v1.34 complex Domcke wave mixing method redone, with redobra instead of conjgpropflag, using action 29 and mcscfnum=1. v1.31 gaugefluxflag and strongcatflag available for photoionization actions 16 17 27 28. With strong fields, recommend explicit back-propagation of cation states with strongcatflag. With ionization during the pulse, recommend (a) velocity gauge calculation or (b) length gauge calculation with gaugefluxflag.eq.2. Either way, photoionization is calculated in velocity gauge, and what remains is only the time-dependent Volkov phase for exact electron KE distribution with action 17 (to-do) v1.31 windowing functions for fourier transform set for each action separately, namelist &parinp variables fttriwindow(iaction) and ftwindowpower(iaction); defaults should be best. Linear damping on [0,t], cosine-squared on [-t:t] v1.28 photoionization during calculation using actions 27 and 28 like actions 16 and 17. Integrals dt only, no energy resolution v1.28 basic integrals added for photoionization actions 16 and 17: integrals dt and domega. New output files, &parinp namelist variables fluxtsumfile (Dat/fluxtsum.dat), projfluxtsumfile, angprojfluxtsumfile give integrated outgoing flux from integrals dt in column 2. spifile and projspifile (xsec.spi.dat and e.g. proj.xsec.spi_001_001.dat) contain outgoing flux integrated domega in column 7 (should agree with integral dt). Integrated outgoing flux corresponds to the change in norm-squared due to ionization. v1.27 best attempt dissociative excitation actions 16,17 with nucfluxopt.eq.2 -- replaces old (version 0) actions 13,14 v1.27 predictor/corrector algorithm modified; results will be different, depending on stepsize par_timestep v1.27 work done by or on each pulse in Hartree output in various files for action 21 emission/absorption. Circular polarization output added action 21 -- new &parinp namelist variable act21circ.eq.0 default -- set act21circ.ne.0 to enable circular polarization output v1.26 angular differential partial photoionization, action 17, with angularflag.ne.0 v1.20 multiple cation files for partial photoionization action 17 using numcatfiles, catspffiles, catavectorfiles in namelist &parinp cation.configlist.BIN no longer needed v1.19 better improvedquadflag=2 and 3 (orbitals). Excited states may be calculated more easily. v1.18 predictor/corrector algorithm modified; results will be different, depending on stepsize par_timestep v1.16 FACTOR OF 1/3 REMOVED FROM PHOTOIONIZATION (Actions 16,17) **** photoionization RESULTS WILL BE 3x GREATER NOW - BEWARE **** the factor of 1/3 is historical. The cross section now reported in column 3 of xsec.spi.dat, xsec.proj.spi.datXXXX, is the quantum cross section in Megabarns for the problem at hand. Perform a 1:2 weighted average of parallel and perpendicular to obtain the cross section averaged over orientations The ionization and absorption cross section for H2 is about 12 Mb at ion onset for both perpendicular and parallel. Not 4. He xsec is about 8. ALSO v1.16 new windowing function for fourier transforms implemented further updates for v1.31 (see above) ALSO v1.16 ABSOLUTE UNITS (megabarns) IMPLEMENTED FOR ACTION 21 emission/absorption. Column 9 in ZDipoleft.Dat etc is cross section in Mb. See H2 and Helium examples. ALSO v1.16 par_consplit.ne.0 attempted first time. Total MPI parallelization of LBNL-AMO-MCTDHF calculation with par_consplit.ne.0 and parorbsplit.eq.3 (parorbsplit.eq.3 for polyatomic only) CHANGES TO PULSE FOURIER TRANSFORM SUBROUTINES IN VERSION 1.10. VARIABLES EFLUXLO, EFLUXHI, NEFLUX, FLUXNBINS, FLUXSINEOPT DEPRECATED. NOTE NEW VARIABLE pulseft_estep USED FOR Pulseft.Dat OUTPUT ONLY ----------------------------------------------------------------------- Please watch the commits to stay notified about bugs and bug fixes: https://github.com/LBNL-AMO-MCTDHF/V1/commits/master Known bugs appear as issues: https://github.com/LBNL-AMO-MCTDHF/V1/issues ----------------------------------------------------------------------- The scripts in the example directories use bash and gnuplot. Using the bash shell is recommended but may not be necessary. To begin a bash shell, if it is not your default, simply execute the command prompt> bash To set bash as your default shell on mac or other linux, execute the command prompt> chsh -s /bin/bash Lots of scripts use gnuplot. Gnuplot should be installed on your system, if you want to have an easy time seeing the results of the example calculations. If gnuplot is installed on your system then the command prompt> which gnuplot should return the location of the executable file. If it returns nothing then it is not installed. The best way to install gnuplot is probably with macports. Possible workflow described below for mac. The general idea is that you can work entirely within the directory that you clone and to which you pull, not copy the git distribution elsewhere for compilation. ---------------------------------------------------------------------------- On Mac: minimal demonstration. *BIN*mac* directories use GFORTRAN. You must have gfortran installed to use the *BIN*mac* directories. To install gfortran -- which you will need to do if the Makeme step fails -- visit https://gcc.gnu.org/wiki/GFortranBinariesMacOS 00) mkdir V1 cd V1 git clone https://github.com/LBNL-AMO-MCTDHF/V1 . or download using a web browser https://github.com/LBNL-AMO-MCTDHF/V1/archive/master.zip and unzip this file unzip master.zip mv master V1 cd V1 PERFORM ALL OTHER STEPS IN THIS V1 DIRECTORY. 0) You must get the examples now separately. cd EXAMPLES-DEPOT rm README git clone https://github.com/LBNL-AMO-MCTDHF/examples-depot . or download using a web browser https://github.com/LBNL-AMO-MCTDHF/EXAMPLES-DEPOT/archive/master.zip and unzip this file to get the EXAMPLES-DEPOT examples files. 1) cd COMPDIRS/BIN.ecs.hermnorm.mac ./Makeme if this fails, you need gfortran. Also you may then cd ../debug.BIN.ecs.hermnorm.mac ./Makeme cd ../BIN.ecs.hermnorm.mac.mpi ./Makeme cd ../debug.BIN.ecs.hermnorm.mac.mpi ./Makeme [ You can delete the directories in COMPDIRS that you don't want, e.g. cd COMPDIRS rm -r *BIN*law* *BIN*ediso* *BIN*cori* ] 2) You need to have the code in your path. Safe, temporary way: 2A) cd COMPDIRS/BIN.ecs.hermorm.mac export PATH=$PATH:$PWD 2B) Permanent way, system-wide, links in /opt/local/bin, requires root password: cd COMPDIRS/BIN.ecs.hermorm.mac ./LinkMe cd ../debug.BIN.ecs.hermorm.mac ./LinkMe debug cd ../BIN.ecs.hermorm.mac.mpi ./LinkMe mpi cd ../debug.BIN.ecs.hermorm.mac.mpi ./LinkMe mpi.debug 2C) Permanent way, user-specific, script that edits your .bashrc file cd COMPDIRS/BIN.ecs.hermnorm.mac ./LinkMeLocal You must restart the terminal after step 2C. 3) mkdir EXAMPLES cp -R -p EXAMPLES-DEPOT/H2-EXAMPLE-PLAIN-SCRIPTS EXAMPLES 4) cd EXAMPLES/H2-EXAMPLE-PLAIN-SCRIPTS follow the README; ./Relax.Bat ./Fourier.Bat ./Flux.Bat 500 ./gnu.xsec for total photoionization; see the README for more. If you want to update the code then perform steps 5 and 6, like steps 1-4, in the V1 working directory, 5) git pull https://github.com/LBNL-AMO-MCTDHF/V1 git pull --tags https://github.com/LBNL-AMO-MCTDHF/V1 then ./Makeme should be sufficient, in the BIN directories, but ./Makeme clean; ./Makeme would certainly be fail safe 6) cd EXAMPLES-DEPOT git pull https://github.com/LBNL-AMO-MCTDHF/examples-depot git pull --tags https://github.com/LBNL-AMO-MCTDHF/examples-depot If you want to switch versions of the code then do e.g. git checkout tags/v1.21 cd COMPDIRS for dir in *BIN*; do cd $dir; ./Makeme clean; ./Makeme & cd ..; done This should work for tags v1.16 and later. Return to the most up-to-date version of the code by executing git checkout master and again recompiling. ---------------------------- MODULES TO LOAD ----------------------------- ---------------------------- ON NERSC & LAWRENCIUM ----------------------------- COMPDIRS/BIN.ecs.hermnorm.law, COMPDIRS/BIN.ecs.hermnorm.edison, etc. LAWRENCIUM: module load mkl EDISON: module unload cray-libsci module load mkl
About
LBNL-AMO-MCTDHF v1.36! . . . . .
Resources
License
Stars
Watchers
Forks
Packages 0
No packages published