the setting of ocp_set for multi-k-points system in tddft.

[Jweijian]

Details

I am trying to use abacus to do tddft calculation. According to the tutorials, i must set ocp for different kpoint and band, such as for 70 bands and 132 electrons system which KPT sets 2 2 1 0 0 0, i set ocp_set 652 21 30 652 21 30 652 21 30 652 21 30. But, when i set ocp_set for multi_kpoints, a warning always raises.
Below is the INPUT and warning information:

Have you read FAQ on the online manual http://abacus.deepmodeling.com/en/latest/community/faq.html

• Yes, I have read the FAQ part on online manual.

Task list for Issue attackers (only for developers)

• Understand the problem or question described by the user.
• Check if the issue is a known problem or has been addressed in the documentation.
• Test the issue or problem on a similar system or environment, if possible.
• Identify the root cause or provide clarification on the user's question.
• Provide a step-by-step guide, including any necessary resources, to resolve the issue or answer the question.
• If the issue is related to documentation, update the documentation to prevent future confusion (optional).
• If the issue is related to code, consider implementing a fix or improvement (optional).
• Review and incorporate any relevant feedback from users or developers.
• Ensure the user's issue is resolved or their question is answered and close the ticket.

[AsTonyshment]

Do you have complete INPUT, KPT, and STRU files for us to debug? @Jweijian

[Jweijian]

Do you have complete INPUT, KPT, and STRU files for us to debug? @Jweijian

I have uploaded the input file in this address
https://github.com/Jweijian/Abacus_debug.git

[AsTonyshment]

I noticed that you have closed this issue. Could you confirm if the problem has been resolved?😂 @Jweijian
If not, you can reopen this issue.

[Jweijian]

I noticed that you have closed this issue. Could you confirm if the problem has been resolved?😂 @Jweijian If not, you can reopen this issue.

Sorry, my bad. I have reopened this issue.

[AsTonyshment]

@Jweijian In your example of CsPbI₃, since you have multiple $\boldsymbol{k}$-points (4 in this case), the average occupation should be 0.5 for each $\boldsymbol{k}$-point. Given that there are a total of 592 electrons, they occupy 296 bands (with nspin=1), with an occupation of 0.5 per $\boldsymbol{k}$-point.

If you want to excite a single electron from HOMO to LUMO, you might set ocp_set as follows:
295*0.5 2*0.25 3*0 295*0.5 2*0.25 3*0 295*0.5 2*0.25 3*0 295*0.5 2*0.25 3*0 (instead of 295*2 2*1 3*0 295*2 2*1 3*0 295*2 2*1 3*0 295*2 2*1 3*0). Otherwise, the total occupation number will exceed the total number of electrons in the system when checking the electron count, resulting in an error.

Useful information: The sequence of $\boldsymbol{k}$-points corresponds to those listed in the output file kpoints. The source code responsible for setting occupations is shown below. Here, the vector ocp_kb is parsed from the string ocp_set, and wg(ik, ib) represents the actual weight for $\boldsymbol{k}$-point ik (a total of wg.nr=4 in this case) and band ib (a total of wg.nc=300 in this case) used in the code for calculations:

abacus-develop/source/module_elecstate/elecstate.cpp

Lines 44 to 50 in 105fd19

	for (int ik = 0; ik < this->wg.nr; ++ik)
	{
	for (int ib = 0; ib < this->wg.nc; ++ib)
	{
	this->wg(ik, ib) = ocp_kb[ik * this->wg.nc + ib];
	}
	}

If you have further questions, feel free to ask! This issue might be due to an unclear description in the user manual. I will create a Pull Request to update the description of ocp_set.

[Jweijian]

@Jweijian In your example of CsPbI₃, since you have multiple k -points (4 in this case), the average occupation should be 0.5 for each k -point. Given that there are a total of 592 electrons, they occupy 296 bands (with nspin=1), with an occupation of 0.5 per k -point.

If you want to excite a single electron from HOMO to LUMO, you might set ocp_set as follows: 295*0.5 2*0.25 3*0 295*0.5 2*0.25 3*0 295*0.5 2*0.25 3*0 295*0.5 2*0.25 3*0 (instead of 295*2 2*1 3*0 295*2 2*1 3*0 295*2 2*1 3*0 295*2 2*1 3*0). Otherwise, the total occupation number will exceed the total number of electrons in the system when checking the electron count, resulting in an error.

Useful information: The sequence of k -points corresponds to those listed in the output file kpoints. The source code responsible for setting occupations is shown below. Here, the vector ocp_kb is parsed from the string ocp_set, and wg(ik, ib) represents the actual weight for k -point ik (a total of wg.nr=4 in this case) and band ib (a total of wg.nc=300 in this case) used in the code for calculations:

abacus-develop/source/module_elecstate/elecstate.cpp

Lines 44 to 50 in 105fd19

for (int ik = 0; ik < this->wg.nr; ++ik)
{
for (int ib = 0; ib < this->wg.nc; ++ib)
{
this->wg(ik, ib) = ocp_kb[ik * this->wg.nc + ib];
}
}
If you have further questions, feel free to ask! This issue might be due to an unclear description in the user manual. I will create a Pull Request to update the description of ocp_set.

Thank you for your kind answers, your suggestions are valid