2_training/forces/phonon_slides.py

#!/usr/bin/env python
"""Phonon bar chart for slides - APS style.
Uses pre-computed data: _phonopy_freqs_ref.json and reference qpoints.yaml.
Outputs: pictures_ml/nequip_phonons.png
"""
import json
import os
import re

import numpy as np
import yaml
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
OUT_PATH   = '/home/apolyukhin/git/aps_slides/random_slides/pictures_ml/nequip_phonons.png'

REF_QE_QPOINTS_YAML = (
    '/home/apolyukhin/scripts/ml/diamond-qe/diamond_epc/displacements/qpoints.yaml'
)

# =============================================================================
# APS slides style
# =============================================================================
marp_text_color = "#575279"
color_dfpt = "mediumseagreen"   # DFT color
color_ml   = "black"            # ML color
alpha        = 0.8
legend_alpha = 0.5
line_width   = 3
fontsize     = 22

plt.rcParams.update({
    'font.size':        fontsize,
    'mathtext.fontset': 'cm',
    'text.color':       marp_text_color,
    'axes.labelcolor':  marp_text_color,
    'xtick.color':      marp_text_color,
    'ytick.color':      marp_text_color,
    'axes.edgecolor':   marp_text_color,
    'axes.labelpad':    10,
})

cm1_per_thz = 33.35641

# =============================================================================
# Load data
# =============================================================================
freqs_json = os.path.join(SCRIPT_DIR, '_phonopy_freqs_ref.json')
with open(freqs_json) as f:
    ml_thz = np.array(json.load(f)['freqs_thz'])

with open(REF_QE_QPOINTS_YAML) as f:
    data = yaml.safe_load(f)
qe_cm1 = np.array([b['frequency'] for b in data['phonon'][0]['band']])
qe_thz = qe_cm1 / cm1_per_thz

dfpt_cm1 = qe_thz * cm1_per_thz
ml_cm1   = ml_thz * cm1_per_thz

n = max(len(dfpt_cm1), len(ml_cm1))
modes = np.arange(1, n + 1)
d = np.pad(dfpt_cm1, (0, n - len(dfpt_cm1)))
m = np.pad(ml_cm1,   (0, n - len(ml_cm1)))

# MAE
mae = np.mean(np.abs(d - m))
print(f'MAE: {mae:.1f} cm-1')

# =============================================================================
# Plot
# =============================================================================
fig, ax = plt.subplots(figsize=(8, 5.5), facecolor='none')
ax.set_facecolor('none')

w = 0.35
ax.bar(modes - w/2, d, w, label='DFPT', color=color_dfpt, alpha=alpha)
ax.bar(modes + w/2, m, w, label='ML (NequIP)', color=color_ml, alpha=alpha)

ax.set_xlabel('Mode')
ax.set_ylabel('Frequency (cm$^{-1}$)')
ax.set_xticks(modes)
ax.legend(loc='upper left', framealpha=legend_alpha, fontsize=0.75*fontsize)

plt.tight_layout()
plt.savefig(OUT_PATH, dpi=300, transparent=True, bbox_inches='tight')
print(f'Saved: {OUT_PATH}')