Okuyama
okymm****@gmail*****
2013年 1月 8日 (火) 00:01:33 JST
Dear Mr. Fujii,
This is Okuyama.
Thanks for reviewing my patch files.
I read your E-mail. In your E-mail,
I think there is a problem in the source code that you suggested.
You suggested new source code:
742 int InputParser::ParseConditionsCIS(vector<string>*
inputTerms, int parseIndex) const{
743 Parameters::GetInstance()->SetRequiresCIS(true);
744 parseIndex++;
.
.
.
.
827 // unpaired electron population
828 if((*inputTerms)[parseIndex].compare(this->stringCISUnpairedPop)
== 0){
829 if((*inputTerms)[parseIndex+1].compare(this->stringYES) == 0){
830 Parameters::GetInstance()->SetRequiresUnpairedPopCIS(true);
831 }
832 else{
833 Parameters::GetInstance()->SetRequiresUnpairedPopCIS(false);
834 }
835 parseIndex++;
836 }
837 parseIndex++;
838 }
839 return parseIndex;
840 }
I implement the above source code into my source code and
run the implemented source code in my PC.
In CIS input, the following input is OK:
unpaired_electron_population yes
or
unpaired_electron_population no
However, the calculation stops abnormaly (segmentation fault)
when the following input:
unpaired_electron_population
Now, I'm investigating the problem.
I may have any time to solve this problem since I can't solve the
problem....
I think I'd like to have a chance to talk with you by mobile phone etc.
Best regards.
Okuyama
On 2013年01月06日 02:55, Mikiya Fujii wrote:
> Dear Okuyama,
>
> Thanks for your patches. I had reviewed them.
>
> ***
> Your coding-grammar and -style are quite fine!
> But, I ask you to fix only one modification about input control:
> (1) UEP should be set as "unpaired_electron_population yes" in CIS-directive,
> that is, the following input obeys the MolDS's input style.
>
> CIS
> ....
> mulliken m
> mulliken n
> unpaired_electron_population yes
> ....
> END
>
> The corresponding code in the InputParser.cpp (r1234 which your 2
> patches and my modification had been applied to) is written as
>
> 742 int InputParser::ParseConditionsCIS(vector<string>*
> inputTerms, int parseIndex) const{
> 743 Parameters::GetInstance()->SetRequiresCIS(true);
> 744 parseIndex++;
> .
> .
> .
> .
> 827 // unpaired electron population
> 828 if((*inputTerms)[parseIndex].compare(this->stringCISUnpairedPop)
> == 0){
> 829 if((*inputTerms)[parseIndex+1].compare(this->stringYES) == 0){
> 830 Parameters::GetInstance()->SetRequiresUnpairedPopCIS(true);
> 831 }
> 832 else{
> 833 Parameters::GetInstance()->SetRequiresUnpairedPopCIS(false);
> 834 }
> 835 parseIndex++;
> 836 }
> 837 parseIndex++;
> 838 }
> 839 return parseIndex;
> 840 }
>
> With integration of the (1) to your second patch "add_opt_ctrl.patch",
> please commit your code to the trunk.
>
>
> ***
> Besides, I ask you to do two more tasks about this ticket.
> [1] Update test input and output files in the molds/trunk/test for the UEP.
> I think that updating of only "c2h6_pm3_directCIS_singlet.in" and
> "c2h6_pm3_directCIS_singlet.dat" is enough.
> [2] Update documents (molds/trunk/doc/README.txt) to use UEP.
> Explanation of the UEP option should be written in the
> <CIS>-paragraph of the README.txt.
>
> Of course, please commit [1] and [2] too.
>
>
> Please ask me if you have any question.
> 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/
> ***************************************
>
>
> 2013/1/4 Okuyama <okymm****@gmail*****>:
>> Dear colleagues
>>
>> This is Okuyama.
>>
>> I send patch files to impliment the unpaired electron population (UEP)
>> and the input control to this mailing list.
>> The patch files are the following files:
>> *calc_uep_without_opt_ctrl.patch
>> *add_opt_ctrl.patch
>>
>> calc_uep_without_opt_ctrl.patch adds source code to be available the
>> unpaired electron population.
>> By appling this patch file, it's possible to calculate the unpaired
>> electron population of excited states on each single atom in molecule.
>> Upto this implimentation, you can calculate the unpaired electron
>> population of nth-excited states
>> by adding the follwing input:
>> CIS
>> mulliken n
>> END
>> Next, the file, add_opt_ctrl.patch adds the input control of the
>> unpaired electron population.
>> The input control executes the calculation of the UEP
>> when there is a keyword, 'unpaired_electron_population' in the CIS input.
>>
>> *Procedure to apply the patch files to the MolDS
>> Revision of the Molds is 1234
>> In a directory, trunk,
>> 1. patch -p0 < calc_uep_without_opt_ctrl.patch
>> 2. patch -p2 < add_opt_ctrl.patch
>>
>> *How to use the UEP calculation
>> Example) Calculation of the UEP of the m and n-th excited states
>> CIS
>> ....
>> mulliken m
>> mulliken n
>> unpaired_electron_population
>> ....
>> END
>>
>> I have made sure the following terms:
>> * There are not segmentation fault and abort by executing all the
>> test inputs.
>> * The MolDS.out is created by executing make depend INTEL=64 &&
>> make INTEL=64
>> * For all the test inputs including CIS keyword, I added the
>> follwing keyword:
>> CIS
>> ....
>> mulliken 2
>> mulliken 3
>> unpaired_electron_population
>> ...
>> END
>> Under above condition, I make sure that there is no segmentation
>> fault.
>>
>> At last, in this implimentation, it will not calculate the UEP at the
>> ground state.
>> The reason why the UEP at the ground state is zero and, therefore this
>> calculation is nonsense.
>>
>> Best regards.
>>
>> -------------- next part --------------
>> テキスト形式以外の添付ファイルを保管しました...
>> ファイル名: calc_uep_without_opt_ctrl.patch
>> 型: text/x-patch
>> サイズ: 8604 バイト
>> 説明: 無し
>> 
Download
>> -------------- next part --------------
>> テキスト形式以外の添付ファイルを保管しました...
>> ファイル名: add_opt_ctrl.patch
>> 型: text/x-patch
>> サイズ: 4357 バイト
>> 説明: 無し
>> URL: http://lists.sourceforge.jp/mailman/archives/molds-dev/attachments/20130104/ce956b15/attachment-0001.bin
>> _______________________________________________
>> MolDS-dev mailing list
>> MolDS****@lists*****
>> http://lists.sourceforge.jp/mailman/listinfo/molds-dev