Multiwfn forum

Manceau

Member

Registered: 2020-07-04

Posts: 4

Inputting error with the "cond" command

Hi,
I want to plot the PDOS of the valence states for a particular atom, such as the p state of Fe.
I followed the instruction on page 498 of the tutorial, and typed the command "cond" in the fragment definition interface, but got the following error:
 

What am I doing wrong?

Thank you,

Alain

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,468

Website

Re: Inputting error with the "cond" command

You should first input a fragment index (e.g. 1) to enter the interface of defining the fragment, then in that interface you can use "cond" command to add specific objects into the current fragment.

Manceau

Member

Registered: 2020-07-04

Posts: 4

Re: Inputting error with the "cond" command

Thank you for your quick answer.
Yes, this option shows up under this option:
7 Set the method for calculating PDOS, current: Mulliken

However, if I choose
7 Set the method for calculating PDOS, current: Becke
it does not appear in the fragment definition menu

and when I define the fragments with the Mulliken option and afterward decide to calculate the PDOS with the Becke method, there is this prompt:
Fragment definitions will be cleaned, OK? (y/n)

Can I use the Mulliken analysis method to construct the inter-fragment composition, and afterward keep this composition whatever the method used to calculate the PDOS (Becke,  Hirshfeld)? Otherwise, is it physically meaningful to answer 'no" to the question?

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,468

Website

Re: Inputting error with the "cond" command

The "cond" command is used to select basis functions according to filtering conditions, however in the case of Becke or Hirshfeld, the fragment is defined as a set of atoms rather than a set of basis functions, therefore there is no "cond" command.

In principle, it is acceptable to plot OPDOS based on Mulliken method while obtain PDOS based on Becke or Hirshfeld method.

Manceau

Member

Registered: 2020-07-04

Posts: 4

Re: Inputting error with the "cond" command

Thank you for the clarification. However, I don't see how to plot, say, the s states of Fe, using the Becke method if only the Fe atom is chosen.

On the left graph, I choose Fe and S states with cond under the Mulliken method. On the right graph, I selected the Becke method and kept the Mulliken fragment definition. I mean, answered "no" to the question " Fragment definitions will be cleaned, OK? (y/n)"

Which PDOS do you recommand, or is there another option which I missed? Thank you.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,468

Website

Re: Inputting error with the "cond" command

The DOS curves generated by different orbital composition calculation methods cannot be shown together in the same map in Multiwfn. You should use "3 Export curve and line data to plain text file in current folder" to export curve data as .txt file based on Mulliken, and then export the curve based on Becke, and finally use third-part tool such as Origin to plot them together.

Manceau

Member

Registered: 2020-07-04

Posts: 4

Re: Inputting error with the "cond" command

Yes, I know this. It was not my question.