phonopy/scripts/pdosplot

202 lines
6.6 KiB
Python
Executable File

#!/usr/bin/env python
# Copyright (C) 2011 Atsushi Togo
# All rights reserved.
#
# This file is part of phonopy.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
#
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
#
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in
# the documentation and/or other materials provided with the
# distribution.
#
# * Neither the name of the phonopy project nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
# FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
# COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
# BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
# CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
# ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
# POSSIBILITY OF SUCH DAMAGE.
# PDOS plot (pdosplot)
#
# Usage:
# pdosplot -i "1 2, 4 5" -o "pdos.pdf"
#
# The axis resolved PDOS is summed up with the successive
# indices separated by ",". In this example, indices 1 and
# 2, 3 and 4 are summed respectively, and then they are
# ploted respectively.
#
# The indices are defined like:
# 1 2 3 : X Y Z of the 1st atom,
# 4 5 6 : X Y Z of the 2nd atom,
# ...
import numpy as np
import matplotlib.pyplot as plt
# Parse options
from optparse import OptionParser
parser = OptionParser()
parser.set_defaults( output_filename = None,
factor = 1.0,
legend_labels = None,
xlabel = None,
ylabel = None,
show_legend = False,
pdos_indices = None,
ymax = None,
ymin = None,
title = None,
f_max = None,
f_min = None )
parser.add_option("--factor", dest="factor", type="float",
help="Factor is multiplied with DOS.")
parser.add_option("-l", "--legend", dest="show_legend",
action="store_true",
help="Show legend")
parser.add_option("--legend_labels", dest="legend_labels",
action="store", type="string",
help="Set legend labels")
parser.add_option("--xlabel", dest="xlabel",
action="store", type="string",
help="Set x label")
parser.add_option("--ylabel", dest="ylabel",
action="store", type="string",
help="Set y label")
parser.add_option("-i", "--indices", dest="pdos_indices",
action="store", type="string",
help="Indices like 1 2, 3 4 5 6...")
parser.add_option("-o", "--output", dest="output_filename",
action="store", type="string",
help="Output filename")
parser.add_option("-t", "--title", dest="title",
action="store", type="string",
help="Title of plot")
parser.add_option("--ymax", dest="ymax", type="float",
help="Maximum value of y axis")
parser.add_option("--ymin", dest="ymin", type="float",
help="Minimum value of y axis")
parser.add_option("--fmax", dest="f_max", type="float",
help="Maximum frequency plotted")
parser.add_option("--fmin", dest="f_min", type="float",
help="Minimum frequency plotted")
(options, args) = parser.parse_args()
# Read data file
frequencies = []
dos = []
filename = "partial_dos.dat"
if len(args) > 0 :
filename = args[0]
for line in open(filename):
if line.strip().split()[0] == '#' or line.strip().split() == '':
continue
tmp_array = [float(x) for x in line.split()]
frequencies.append(tmp_array.pop(0))
dos.append(tmp_array)
frequencies = np.array(frequencies)
dos = np.array(dos).transpose()
# Extract indices
indices = []
if options.pdos_indices is None:
indices.append(range(1, dos.shape[0] + 1))
else:
for v in options.pdos_indices.split(','):
indices.append([int(x) for x in v.split()])
# Set plot range in frequency axis
if options.f_max is None:
max_freq = max(frequencies)
else:
max_freq = options.f_max
if options.f_min is None:
min_freq = min(frequencies)
else:
min_freq = options.f_min
min_i = 0
max_i = len(frequencies)
for i, f in enumerate(frequencies):
if f > max_freq + (frequencies[1] - frequencies[0]) / 10:
max_i = i + 1
break
for i, f in enumerate(frequencies):
if f > min_freq - (frequencies[1] - frequencies[0]) / 10:
min_i = i
break
# Plot
plots = []
for nums in indices:
pdos = np.zeros(frequencies.shape[0], dtype='double')
for v in nums:
pdos += dos[v - 1]
plots.append(plt.plot(
frequencies[min_i:max_i], pdos[min_i:max_i] * options.factor))
plt.grid(True)
plt.xlim(min_freq, max_freq)
if (options.ymin is not None) and (options.ymax is not None):
plt.ylim(options.ymin, options.ymax)
elif (options.ymin is not None) and (options.ymax is not None):
plt.ylim(ymax=options.ymax)
elif (options.ymin is not None) and (options.ymax is None):
plt.ylim(ymin=options.ymin)
if options.xlabel is None:
plt.xlabel('Frequency')
else:
plt.xlabel(options.xlabel)
if options.ylabel is None:
plt.ylabel('Partial density of states')
else:
plt.ylabel(options.ylabel)
if options.show_legend:
if options.legend_labels is not None:
if len(options.legend_labels.split()) == len(plots):
labels = options.legend_labels.split()
else:
print "Number of labels is not same as number of plots."
labels = indices
else:
labels = indices
plt.legend(labels, loc='upper left')
if options.title is not None:
plt.title(options.title)
if options.output_filename is not None:
plt.rcParams['pdf.fonttype'] = 42
plt.rcParams['font.family'] = 'serif'
plt.savefig(options.output_filename)
else:
plt.show()