Mikiya Fujii
mikiy****@gmail*****
2012年 11月 12日 (月) 11:36:27 JST
Dear M. Okuyama, Sorry to late. Last weekend, I had been to Izu by the laboratory travel. > In the MolDS, is the calculation including electron correlation only the > CI ? YES. In this stage, we have a plan to implement SCF and CIS only. > I guess the SCF method in the MolDS does not include electron correlation. > So, the UEP analysis needs to be used at methods without the SCF. YES. SCF in MolDS does not include any electron correlation. Therefore, it seems better that UEP is implemented in CIS-module (i.e. ZndoS::CalcCISProperties() and ZndoS::OutputCISResults()) than SCF. To implement UEP as one of the additional option analyses, "exciton_energies" which is one of the CIS-options may be of best reference. Beside, I think there are no problem to made "atomicElectronPopulation" be a class-variable of Cndo2 like as "Cndo2::freeExcitonEnergiesCIS". But, take care the timing of Malloc and Free of "atomicElectronPopulation". The same treatments with Malloc and Free of "Cndo2::freeExcitonEnergiesCIS" in "ZindoS" are good for UEP also. My best, Mikia *************************************** Mikiya Fujii, Ph.D. Yamashita & Ushiyama Laboratory, Department of Chemical System Engineering, Faculty of Engineering, The University of Tokyo. tel&fax: +81-3-5841-7286 email: fujii****@tcl***** HP: http://www.tcl.t.u-tokyo.ac.jp/ *************************************** 2012/11/9 okuyama <okymm****@gmail*****>: > > Dear Mr. Fujii > > Thanks for your response and opinions. > I will write my response under your question as follows. > > (2012/11/09 10:00), Mikiya Fujii wrote: >> Dear Michihiro, >> >> The following discussion is based on r1106. >> The term "UEP" means the unpaired electron population. >> >> 0)Basically, your coding is fine! > Thanks ! I'm glad. >> >> 1)Although I do not know details of the UEP analysis, >> I think that the unpaired electron analysis is suitable as option. >> I wanna listen to your opinion. > I think so too. > I think that I'd like to be able to calculate the UEP analysis if user > hope the analysis. > In that case, we need to consider about the detail of new options. > In this point, I'd like to hear your opinions, > where I will also consider about that. >> >> If UEP is one of the default analyses, you should modify the test >> suite in /trunk/test. >> If UEP is one of the option analyses, you should modify the input parser. >> >> 2)Can the UEP analysis be applied to both SCF and CI? > Yes, it can theoretically. However, in the implementation at my patch file, > the UEP analysis can't be applied to CI. > >> >> 3)Is the "Cndo2::atomicElectronPopulation" needed to be a class-variable? >> If this UEP is not re-used except in SCF results, >> the "atomicElectronPopulation" is sufficient as a local variable of >> the "Cndo2::DoSCF" or else. > My answer for this question is nothing. > That's because there are two consideration as follows: > 1) The UEP analysis should be used at a system involved electron > correlation. > The reason why the UEP becomes zero in the SCF > as the Hartree-Fock method not included electron correlation. > I guess the SCF method in the MolDS does not include electron correlation. > So, the UEP analysis needs to be used at methods without the SCF. > In the MolDS, is the calculation including electron correlation only the > CI ? > 2) I can't understand deeply a concept of class at the C++ yet. > > In this point, I probably need to have a chance to do a more discussion > by using the E-mail etc. >> >> 4)Line 1599 in Cndo2.cpp is not needed >> because the "Cndo2::atomicElectronPopulation" is initialized at the line 1593. >> I think the line 1580 guided you to miss-coding. >> Sorry, the line 1580 is my mistake, i.e., the 1580 is also not needed. >> I will fix to delete the line 1580. > O.K. > > Best regards. > > M. Okuyama. >> >> My best, >> Mikiya >> *************************************** >> Mikiya Fujii, Ph.D. >> >> Yamashita & Ushiyama Laboratory, >> Department of Chemical System Engineering, >> Faculty of Engineering, >> The University of Tokyo. >> >> tel&fax: +81-3-5841-7286 >> email: fujii****@tcl***** >> HP: http://www.tcl.t.u-tokyo.ac.jp/ >> *************************************** >> >> >> 2012/11/9 okuyama <okymm****@gmail*****>: >>> Dear Mr. Fujii >>> >>> Thanks for your E-mail. >>> O.K. >>> I thought that I'd like to confirm whether I could send the mail or not. >>> That's becauseI start using a new e-mailer. >>> >>> I'min holiday since last Saturday. >>> I'm going to start my job from next Monday. >>> So, I can take enough time to developthe Molds until it. >>> >>> I'm going to take abreakfast now. >>> >>> Best regards. >>> >>> M.Okuyama >>> >>> (2012/11/09 9:25), Mikiya Fujii wrote: >>>> Dear Michihiro, >>>> >>>> I received the mail with the attached files. >>>> Sorry, I could not check your coding yesterday. >>>> >>>> But I am checking your code just now! >>>> Please wait a moment. >>>> >>>> Anyway, are you in holiday? >>>> >>>> Best, >>>> Mikiya >>>> >>>> >>>> >>>> *************************************** >>>> Mikiya Fujii, Ph.D. >>>> >>>> Yamashita & Ushiyama Laboratory, >>>> Department of Chemical System Engineering, >>>> Faculty of Engineering, >>>> The University of Tokyo. >>>> >>>> tel&fax: +81-3-5841-7286 >>>> email: fujii****@tcl***** >>>> HP: http://www.tcl.t.u-tokyo.ac.jp/ >>>> *************************************** >>>> >>>> >>>> 2012/11/9 Okuyama <okymm****@gmail*****>: >>>>> Dear Mr. Fujii >>>>> >>>>> After receiving the mail of the following contents, >>>>> I send the mail with attached files again to Molds-dev mailing list. >>>>> Could you receive my mail ? >>>>> >>>>> Best regards. >>>>> >>>>> M. Okuyama >>>>> >>>>> (2012年11月08日 12:35), Mikiya Fujii wrote: >>>>>> Dear M.Okuyama, >>>>>> >>>>>> Thanks for your contributions!!! >>>>>> However, the attached files were discarded by the mailing-list-system. >>>>>> >>>>>> I modified the setting of the mailing-list-system to allow attached files. >>>>>> Thus, please send the mail with the attached files again. >>>>>> >>>>>> My best, >>>>>> Mikiya >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> *************************************** >>>>>> Mikiya Fujii, Ph.D. >>>>>> >>>>>> Yamashita & Ushiyama Laboratory, >>>>>> Department of Chemical System Engineering, >>>>>> Faculty of Engineering, >>>>>> The University of Tokyo. >>>>>> >>>>>> tel&fax: +81-3-5841-7286 >>>>>> email: fujii****@tcl***** >>>>>> HP: http://www.tcl.t.u-tokyo.ac.jp/ >>>>>> *************************************** >>>>>> >>>>>> >>>>>> 2012/11/8 okymmolds <okymm****@gmail*****>: >>>>>>> Dear Fujii >>>>>>> >>>>>>> A patch file to calculate unpaired electron population >>>>>>> on ground state has been developed, where it's impossible to calculate >>>>>>> the population of excited states. orz >>>>>>> I send this patch file to you. >>>>>>> I wold like to get some opinion from you. >>>>>>> >>>>>>> Best regards. >>>>>>> >>>>>>> M.Okuyama >>>>>>> >>>>>>> ------- >>>>>>> How to patch by this patch file: >>>>>>> >>>>>>> In revision 1106 source code, you can introduce >>>>>>> a calculation of unpaired electron population to >>>>>>> your MolDS by the following command, >>>>>>> >>>>>>> mv unapired_pop.patch (work directory)/trunk/src >>>>>>> cd trunk/src >>>>>>> patch -p0 < unpaired_pop.patch >>>>>>> >>>>>>> --------- >>>>>>> About output of unpaired electron population: >>>>>>> For example, in src directory, if you execute the following command, >>>>>>> >>>>>>> ./MolDS.out ../test/h2o_zindos_directCIS_singlet.in, >>>>>>> >>>>>>> you can find the following output, >>>>>>> >>>>>>> | i-th | atom type | Unpaired electron population[a.u.]| >>>>>>> Unpaired electron population: 0 O 2.220446e-15 >>>>>>> Unpaired electron population: 1 H 0.000000e+00 >>>>>>> Unpaired electron population: 2 H -1.110223e-15 >>>>>>> >>>>>>> . This is the output of unpaired electron population. >>>>>>> I will send this log as 'h2o_zindos_directCIS_singlet.log'. >>>>>>> >>>>>>> The above case is an example. >>>>>>> In the job that calculates Mulliken charge, the calculation >>>>>>> of the unpaired electron population is executed. >>>>>>> >>>>>>> -- >>>>>>> okymmolds <okymm****@gmail*****> >>>>>>> _______________________________________________ >>>>>>> MolDS-dev mailing list >>>>>>> MolDS****@lists***** >>>>>>> http://lists.sourceforge.jp/mailman/listinfo/molds-dev >>>>>> _______________________________________________ >>>>>> MolDS-dev mailing list >>>>>> MolDS****@lists***** >>>>>> http://lists.sourceforge.jp/mailman/listinfo/molds-dev >>>>>> >>>>> -- >>>>> ------------------------- >>>>> Michihiro Okuyama >>>>> okymm****@gmail***** >>>>> ------------------------- >>>>> >>>>> _______________________________________________ >>>>> MolDS-dev mailing list >>>>> MolDS****@lists***** >>>>> http://lists.sourceforge.jp/mailman/listinfo/molds-dev >>>> _______________________________________________ >>>> MolDS-dev mailing list >>>> MolDS****@lists***** >>>> http://lists.sourceforge.jp/mailman/listinfo/molds-dev >>> -------------- next part -------------- >>> HTMLの添付ファイルを保管しました... >>>다운로드 >>> _______________________________________________ >>> MolDS-dev mailing list >>> MolDS****@lists***** >>> http://lists.sourceforge.jp/mailman/listinfo/molds-dev >> _______________________________________________ >> MolDS-dev mailing list >> MolDS****@lists***** >> http://lists.sourceforge.jp/mailman/listinfo/molds-dev > > _______________________________________________ > MolDS-dev mailing list > MolDS****@lists***** > http://lists.sourceforge.jp/mailman/listinfo/molds-dev
