1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
|
\documentclass[10pt,a4paper]{scrartcl}
\usepackage{amsmath}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{color}
\usepackage[english]{babel}
\usepackage[
colorlinks=true,
urlcolor=blue,
linkcolor=green
]{hyperref}
% no indentation
\setlength{\parindent}{0pt}
\begin{document}
\section*{MADNESS: CIS Excitation Energies}
The Tutorial was written for the Git commit \textit{f03a51fbe4f5a30d870c28772e5960ae39502da9} which you can checkout before compiling via: \\
\textit{git checkout f03a51fbe4f5a30d870c28772e5960ae39502da9}\\
If you experience any sort of trouble please feel free to contact me at:\\
\href{mailto:jakob.kottmann@hu-berlin.de}{jakob.kottmann@hu-berlin.de}
\section{Built LRCCS}
Detailed information can be found in the MADNESS wiki on Github: \url{https://github.com/m-a-d-n-e-s-s/madness/wiki}
\begin{itemize}
\item Make shure you have Git installed
\item Make shure you have (see also the MADNESS wiki):
\begin{itemize}
\item C++11 Compiler: Clang-3.8 or later (Mac OS), GCC-4.8 or later (Linux)
\item Corresponding MPICH
\item GNU Autotools (autoconf 2.59 or later, automake 1.11 or later)
\end{itemize}
\item Get the MADNESS source Code from Github\\
\textit{git clone {https://github.com/m-a-d-n-e-s-s/madness.git} madness-source}
\item Execute \textit{autogen.sh} in the madness-source directory
\item Configure MADNESSS in your compile directory like this (use the correct paths):\\
\textit{/path-to-source/madness-source/configure 'CC=/usr/bin/clang' 'CXX=/usr/bin/clang++'
'MPICC=/opt/local/bin/mpicc-mpich-clang38' 'MPICXX=/opt/local/bin/mpicxx-mpich-clang38'}
\item Compile the MADNESS libraries in your compile directory\\
\textit{make -C /path-to-madness-compile/madness-compile/ libraries}
\item Compile cis in the MADNESS examples directory\\
\textit{make -C /path-to-madness-compile/madness-compile/src/apps/cis cis}
\end{itemize}
\section{Run LRCCS}
After compiling you can execute \textit{cis} in the \textit{madness-compile/src/apps/cis} directory. For the calculation it is necessary that an input file with the name \textit{input} is present in the directory from which you call \textit{cis}.
\clearpage
\section{Quick Tutorial}
The programm \textit{LRCCS} \cite{mra-cis} will solve the linear-response equations for excitation energies in the CIS approximation via the following steps.
\begin{enumerate}
\item Solve Hartree-Fock (HF) ground state
\item Create guess vectorfunctions $\lbrace x_i \rbrace $ from the HF orbitals $\lbrace \phi_i \rbrace$ via
\[ x_i = f\cdot\phi_i \]
where $f$ is constructed from polynomials (see Keyword section)
\item Iterate the guess vectorfunctions
\item Iterate the lowest guess vectorfunctions further
\item Iterate sequentially untill convergence
\end{enumerate}
Detailed information can be found in \cite{mra-cis}.
\subsection{Input File Structure}
The input file has four main sections.
\paragraph{Ground-State section:} Begins with \textit{dft} and ends with \textit{end}\\
Mandatory Keywords:
\begin{itemize}
\item xc value: \textit{value} is the exchange correlation potential. To solve the CIS equations this has to be always \textit{hf}
\item k value: \textit{value} is the order of polynomials for the MRA representation (usually around 6-9)
\item econv value: \textit{value} is the MRA:threshold for the HF orbitals and at the same time the convergence threshold for the energy
\end{itemize}
Optional Keywords:
\begin{itemize}
\item L value: \textit{value} it the size of the cubic simulation box in atomic units. Each dimension of the box runs rum -L to L (default is 50.0)
\item dconv value: \textit{value} convergence threshold for the orbitals (does not affect MRA threshold)
\item nuclear\_corrfac value: \textit{value} is the nuclear correlation factor which is used \\
(default is \textit{none}, other options are \textit{slater}, \textit{GradientalGaussSlater}, \textit{GaussSlater}, \textit{LinearSlater} and
\textit{Polynomial}). For more information see \cite{nemo-I}.
\item no\_compute or restart: Load the HF equations from the file \textit{restartdata.00000} which has to be present in the same directory. If \textit{restart} is chosen the HF orbitals will be re-iterated
\end{itemize}
\paragraph{Response Section:} Begins with \textit{cc2} and ends with \textit{end}\\
Mandatory Keywords
\begin{itemize}
\item thresh\_3D value: \textit{value} is the MRA threshold for the response vectorfunctions (default is 5)
\end{itemize}
Optional Keywords
\begin{itemize}
\item tda\_econv\_hard value: \textit{value} is the final convergence threshold for the excitation energies
\item tda\_dconv\_hard value: \textit{value} is the final convergence threshold for the response vectorfunctions
\item tda\_excitations value: \textit{value} is the number of excitation energies which shall be calculated
\item tda\_guess\_excitations value: \textit{value} is the number of guess excitations which is iterated in the beginning
\item tda\_iterating\_excitations value: \textit{value} is the number of excitation vectors which are iterated together (change if you have memory problems)
\item freeze value: \textit{value} is the number of frozen core orbitals
\item tda\_guess value: \textit{value} is the guess that is used (default is a big guess from a perturbed fock matrix). Possible entries are: \textit{dipole}, \textit{dipole+}, \textit{quadrupole}, \textit{big\_fock\_3}, \textit{big\_fock\_4}, \textit{c2v}, \textit{c2v\_big} and \textit{custom}).
The polynomials which are used for the specific guesses will be displayed in the output. The \textit{custom} valueword has to be used together with the \textit{exop} valueword.
\item exop value: \textit{value} is a custom polynomial to create a guess vectorfunction from the HF orbitals.
Example: \textit{exop x 1.0 y 2.0, x 3.0 z 1.0 c -2.0} corresponds to $xy^2 + -2.0\cdot x^3z$.
To calculate more than one guess the exop keyword can used more than one time.
Example: The following line would be equivalent with the \textit{tda\_guess dipole+} keyword: \\
\textit{tda\_guess custom} \\
\textit{exop x 1.0}\\
\textit{exop y 1.0}\\
\textit{exop z 1.0}\\
\textit{exop x 2.0, y 2.0, z 2.0}
\end{itemize}
\clearpage
\paragraph{Geometry section:} Begins with \textit{geometry} and ends with \textit{end}. Contains the molecular coordinates in atomic units
\paragraph{Plot section:} Begins with \textit{plot} and ends with \textit{end}. Contains plotting information. Can in principle be empty but should be present.
\subsection{Example Input}
The example input File should also be found in the madness-source/src/apps/cis directory with the name \textit{input}.\\
dft\\
xc hf\\
econv 1.e-5\\
dconv 1.e-4\\
end\\
cc2\\
thresh\_3D 1.e-4\\
tda\_econv\_hard 1.e-4\\
tda\_dconv\_hard 1.e-3\\
tda\_guess dipole+ \\
end\\
geometry\\
he 0.0 0.0 0.0\\
end\\
plot \\
plane x1 x2\\
origin 0.0 0.0 0.0\\
zoom 1.0\\
end
%References
\bibliographystyle{plain}
\bibliography{references}
\end{document}
|
