module 'quippy' has no attribute 'potential_module'

[9Suyoung9]

I am trying to use the GAP model to predict the energy and force of the structure, but unfortunately my python script gives me the following error. I would like to know how to deal with it?This is my python script：
import ase.io
import numpy as np
import matplotlib.pyplot as plt
from quippy.potential import Potential
from quippy.convert import ase_to_quip
from sklearn.metrics import mean_squared_error
import re

读取所有构型

frames = ase.io.read('test.xyz', index=':')

true_energies = []
pred_energies = []
true_forces = []
pred_forces = []

初始化GAP势

try:
gap_pot = Potential(param_str='IP GAP', param_filename='gap_GaO_20220921.xml')
except Exception as e:
raise RuntimeError(f"Failed to load GAP potential: {str(e)}") from e

for atoms in frames:
# 从comment中提取DFT能量
comment = atoms.info.get('comment', '')
energy_match = re.search(r'energy=([-\d.]+)', comment)
if not energy_match:
raise ValueError(f"Energy not found in comment: {comment}")
true_energy = float(energy_match.group(1))

# 获取DFT计算的力
dft_forces = atoms.arrays['forces']

# 转换ASE Atoms为QUIP Atoms对象
quip_atoms = ase_to_quip(atoms)

# 使用GAP势计算
try:
    gap_pot.calc(quip_atoms)
    pred_energy = quip_atoms.energy
    gap_forces = quip_atoms.forces
except Exception as e:
    raise RuntimeError(f"GAP calculation failed: {str(e)}") from e

# 存储结果
true_energies.append(true_energy)
pred_energies.append(pred_energy)
true_forces.append(dft_forces.ravel())
pred_forces.append(gap_forces.ravel())

转换为numpy数组

true_forces = np.concatenate(true_forces)
pred_forces = np.concatenate(pred_forces)

计算RMSE

energy_rmse = np.sqrt(mean_squared_error(true_energies, pred_energies))
force_rmse = np.sqrt(mean_squared_error(true_forces, pred_forces))

print(f"Energy RMSE: {energy_rmse:.4f} eV")
print(f"Force RMSE: {force_rmse:.4f} eV/Å")

绘制对角线图

plt.figure(figsize=(12, 5))

plt.subplot(1, 2, 1)
plt.scatter(true_energies, pred_energies, alpha=0.6)
min_e = min(min(true_energies), min(pred_energies))
max_e = max(max(true_energies), max(pred_energies))
plt.plot([min_e, max_e], [min_e, max_e], 'k--', lw=1)
plt.xlabel('DFT Energy (eV)')
plt.ylabel('GAP Predicted Energy (eV)')
plt.title(f'Energy Comparison (RMSE={energy_rmse:.4f} eV)')

plt.subplot(1, 2, 2)
plt.scatter(true_forces, pred_forces, alpha=0.6, s=2)
min_f = min(min(true_forces), min(pred_forces))
max_f = max(max(true_forces), max(pred_forces))
plt.plot([min_f, max_f], [min_f, max_f], 'k--', lw=1)
plt.xlabel('DFT Force (eV/Å)')
plt.ylabel('GAP Predicted Force (eV/Å)')
plt.title(f'Force Comparison (RMSE={force_rmse:.4f} eV/Å)')

plt.tight_layout()
plt.savefig('gap_vs_dft_comparison.png')
plt.show()

[gabor1]

perhaps you didn't install the "quippy-ase" package? (the name of the package is given in the README)