- CONTENTS
This package contains the evaluation scripts for reproducing the figures and analysis in the paper M. Dolfi, B. Bauer, S. Keller, M. Troyer, "Pair Correlations in Doped Hubbard Ladders" (to appear).
The package is organized as follow:
-- LICENSE_1_0.txt # License file
-- README.md # This readme
-- data_evaluated/ # Folder containing evaluated data in
# text format
-- data_raw/ # Folder containing raw data in hdf5
# format
-- density_correlations.ipynb # (**) IPython notebook for
# density-density correlations
-- download_evaluated_data.sh # Script downloading a preset of
# evaluated data (see below)
-- download_raw_data.sh # Script downloading the raw data
# (see below)
-- energy.ipynb # (**) IPython notebook for the
# energy extrapolation
-- local_density.ipynb # (**) IPython notebook for local
# density plots, as well as density amplidues analsys
-- pair_correlations.ipynb # (**) IPython notebook for pair
# correlations
-- scripts/ # Scripts for data evaluation and
# plotting
-- scripts/amplitudes.py # (**) Performs the density-amplitudes
# analysis
-- scripts/corr_helpers.py # helper functions
-- scripts/density.py # (**) Extracts the density and fit
# the density
-- scripts/density_correlations.py # (**) Extracts the density-density
# correlations
-- scripts/energy.py # (**) Extract the energy
-- scripts/extrapolate.py # (**) Extrapolation functions for the
# energy
-- scripts/extrapolate_local.py # (**) Extrapolation functions for
# vector observables (density and correlation functions)
-- scripts/load.py # (**) Load functions which load the
# evaluated data or perform the data analysis if not yet cached
-- scripts/load_raw_data.py # (*) Load functions which load the measurements
# from the raw data
-- scripts/pairfield_correlations.py # (**) Extracts the pairing
# correlations
-- scripts/pyalps_dset/ # subset of the pyalps library to perform data analysis
# correlations
-- scripts/utils.py # Utilities
(*) requires pyalps in the Python PATH
(**) requires pyalps only if new obserservables have to be extracted from the raw data
- DEPENDENCIES
For reproducing the figures in the paper from the evaluated data:
- Python 2.7
- IPython
- Numpy
- SciPy
- Matplotlib
Additionally for repeating the analysis from the raw data and/or extend the evaluation:
- Pyalps (http://alps.comp-phys.org)
- OBTAIN THE DATA
We have two sets of data:
- Extracted data is stored in
data_extracted/. If the directory is empty, you can download pre-extracted observables from the data DOI.
doi: http://dx.doi.org/10.7910/DVN/I5ANSU
The script download_extracted_data.sh will perform the operation for you.
Extracted data contain the all observables used in the paper in text format. Correlation functions have already been exported in the form C(M, r) with M being the bond dimension and r=|i-j| the distance between two rungs. As described in the paper we average over several pair i,j. Correlation functions for r=|i-j| with a fixed i=18 or i=20 are also exported to reproduce the comparison in the paper.
More observables or other correlation function analysis can be extracted from the raw data (see next part).
Data is stored in text format which can be read with your preferred tool, the initial lines starting with # are inteded as comments.
- Raw data is stored in
data_raw/. If the directory is empty, you need to download the raw data from the the data DOI.
doi: http://dx.doi.org/10.7910/DVN/I5ANSU
The script download_raw_data.sh will perform the operation for you.
Raw data is stored in HDF5 according to the ALPS Schema
(http://alps.comp-phys.org). The loader script in scripts/load_raw_data.py
is an example on how to read the raw data.
In this package we provide data for these parameters:
- Bond dimension M = 800, 1200, 1600, 2000, 2800, 3200, 3600, 4000, 4800
- System size L = 32, 48, 64, 80, 96, 128, 160, 192
- Odd system sizes (only partial data) L = 33, 49, 65, 81, 97, 129
- Averange filling n = 0.875, 0.9375, 0.96875
- START THE WORKFLOWS
Evaluation workflows to reproduce the figures in the paper as available as IPython notebooks and simple Python scripts. In what follows we describe only how to run the notebooks; for the Python scripts the procedure is analogous.
- Launch the notebook from the root of this package:
ipython notebookThis should open a browser window, where you can see the content of the package.
-
Open the notebooks (files ending in .ipynb) by clicking on them.
-
LICENSE
Distributed under the Boost Software License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
We ask that you acknowledge the use of our data and/or analysis workflows by citing the following paper: M. Dolfi, B. Bauer, S. Keller, M. Troyer, "Pair Correlations in Doped Hubbard Ladders" (to appear).