Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userImplementation of the Linear Method for the optimization of Jastrow-Feenberg and Backflow Correlations
102
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
We present a fully detailed and highly performing implementation of the Linear Method [J. Toulouse and C. J. Umrigar (2007)] to optimize Jastrow-Feenberg and Backflow Correlations in many-body wave-functions, which are widely used in condensed matter physics. We show that it is possible to implement such optimization scheme performing analytical derivatives of the wave-function with respect to the variational parameters achieving the best possible complexity O(N^3) in the number of particles N.
rate research
Read More
We introduce an efficient way to improve the accuracy of projected wave functions, widely used to study the two-dimensional Hubbard model. Taking the clue from the backflow contribution, whose relevance has been emphasized for various interacting systems on the continuum, we consider many-body correlations to construct a suitable approximation for the ground state at intermediate and strong couplings. In particular, we study the phase diagram of the frustrated $t{-}t^prime$ Hubbard model on the square lattice and show that, thanks to backflow correlations, an insulating and non-magnetic phase can be stabilized at strong coupling and sufficiently large frustrating ratio $t^prime/t$.
We propose a Jastrow factor for electron-electron correlations that interpolates between the radial symmetry of the Coulomb interaction at short inter-particle distance and the space-group symmetry of the simulation cell at large separation. The proposed Jastrow factor captures comparable levels of the correlation energy to current formalisms, is 40% quicker to evaluate, and offers benefits in ease of use, as we demonstrate in quantum Monte Carlo simulations.
We describe an open-source implementation of the continuous-time interaction-expansion quantum Monte Carlo method for cluster-type impurity models with onsite Coulomb interactions and complex Weiss functions. The code is based on the ALPS libraries.
The so-called local density approximation plus the multi-orbital mean-field Hubbard model (LDA+U) has been implemented within the all-electron projector augmented-wave method (PAW), and then used to compute the insulating antiferromagnetic ground state of NiO and its optical properties. The electronic and optical properties have been investigated as a function of the Coulomb repulsion parameter U. We find that the value obtained from constrained LDA (U=8 eV) is not the best possible choice, whereas an intermediate value (U=5 eV) reproduces the experimental magnetic moment and optical properties satisfactorily. At intermediate U, the nature of the band gap is a mixture of charge transfer and Mott-Hubbard type, and becomes almost purely of the charge-transfer type at higher values of U. This is due to the enhancement of the oxygen 2p states near the top of the valence states with increasing U value.
We present the $texttt{TRIQS}/texttt{SOM}$ analytic continuation package, an efficient implementation of the Stochastic Optimization Method proposed by A. Mishchenko et al [Phys. Rev. B $textbf{62}$, 6317 (2000)]. $texttt{TRIQS}/texttt{SOM}$ strives to provide a high quality open source (distributed under the GNU General Public License version 3) alternative to the more widely adopted Maximum Entropy continuation programs. It supports a variety of analytic continuation problems encountered in the field of computational condensed matter physics. Those problems can be formulated in terms of response functions of imaginary time, Matsubara frequencies or in the Legendre polynomial basis representation. The application is based on the $texttt{TRIQS}$ C++/Python framework, which allows for easy interoperability with $texttt{TRIQS}$-based quantum impurity solvers, electronic band structure codes and visualization tools. Similar to other $texttt{TRIQS}$ packages, it comes with a convenient Python interface.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
