-
Notifications
You must be signed in to change notification settings - Fork 5
SOF-7777 Update: add python code to get VBO polar #265
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
Changes from all commits
4574326
1832d1b
050c460
d1a33cc
e522970
878de8c
f6a7032
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,5 @@ | ||
| numpy<2.0 | ||
| scipy>=1.5.4 | ||
| matplotlib>=3.0.0 | ||
| ase>=3.22.1 | ||
| mat3ra-made[tools]>=2024.11.12.post0 |
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,193 @@ | ||
| # ------------------------------------------------------------------ # | ||
| # Linear Fit of ESP for Polar Interface VBO Calculation # | ||
| # ------------------------------------------------------------------ # | ||
| # | ||
| # Reference: Choudhary & Garrity, arXiv:2401.02021 (InterMat) # | ||
| # # | ||
| # For polar interfaces, ESP shows linear gradient in bulk regions # | ||
| # due to internal electric field. We fit each slab region and use # | ||
| # the average value of the fit as the ESP reference. # | ||
| # # | ||
| # VBO Calculation: # | ||
| # 1. Fit interface profile over slab 1 region → Va_interface # | ||
| # 2. Fit interface profile over slab 2 region → Vb_interface # | ||
| # 3. Fit bulk left profile over slab 1 region → Va_bulk_left # | ||
| # 4. Fit bulk right profile over slab 2 region → Vb_bulk_right # | ||
| # 5. VBO = (∆V_interface) - (∆V_bulk) # | ||
| # where ∆V_interface = Vb_interface - Va_interface # | ||
| # ∆V_bulk = Vb_bulk_right - Va_bulk_left # | ||
| # # | ||
| # Input: # | ||
| # - profile_left, profile_right: ESP profiles for bulk materials # | ||
| # - profile_interface: ESP profile for interface structure # | ||
| # # | ||
| # Output: VBO (Valence Band Offset) # | ||
| # # | ||
| # NEW: Slab boundaries auto-detected using fingerprint matching # | ||
| # ------------------------------------------------------------------ # | ||
| import json | ||
|
|
||
| import matplotlib | ||
| import ase.io | ||
| from mat3ra.made.material import Material | ||
| from mat3ra.made.tools.convert import from_ase | ||
|
|
||
| # Non-interactive backend for running the script on the server, if working in Jupyter, comment out the next line | ||
| matplotlib.use("Agg") | ||
| import matplotlib.pyplot as plt | ||
| import numpy as np | ||
| from types import SimpleNamespace | ||
| from scipy.stats import linregress | ||
|
|
||
| # Read structure from pw_scf.out (generated by previous pw_scf calculation) | ||
| # Material index: 0=Interface, 1=Left, 2=Right | ||
| # Files are named: pw_scf.out, pw_scf.out-1, pw_scf.out-2 | ||
| pw_scf_output = f"./pw_scf.out" | ||
| pw_scf_output_1 = f"./pw_scf.out-1" | ||
| pw_scf_output_2 = f"./pw_scf.out-2" | ||
|
|
||
| # Read atomic structure from espresso output | ||
| atoms = ase.io.read(pw_scf_output, format="espresso-out") | ||
| atoms_1 = ase.io.read(pw_scf_output_1, format="espresso-out") | ||
| atoms_2 = ase.io.read(pw_scf_output_2, format="espresso-out") | ||
|
|
||
| # Convert ASE Atoms to Material using mat3ra-made | ||
| material = Material.create(from_ase(atoms)) | ||
| material_1 = Material.create(from_ase(atoms_1)) | ||
| material_2 = Material.create(from_ase(atoms_2)) | ||
|
|
||
| material.to_cartesian() | ||
| material_1.to_cartesian() | ||
| material_2.to_cartesian() | ||
|
|
||
| # Get the z-coordinate boundaries of each slab using element-based matching | ||
| coords = material.basis.coordinates.values | ||
| elements = material.basis.elements.values | ||
| z_elements = sorted(zip([c[2] for c in coords], elements)) | ||
| n_left = len(material_1.basis.elements.values) | ||
|
|
||
| z_max_1 = z_elements[n_left - 1][0] # Last atom of left slab | ||
| z_min_2 = z_elements[n_left][0] # First atom of right slab | ||
| z_min_1 = z_elements[0][0] | ||
| z_max_2 = z_elements[-1][0] | ||
|
|
||
| print(f"Detected Slab 1 (left) boundaries: z = {z_min_1:.3f} to {z_max_1:.3f} Å") | ||
| print(f"Detected Slab 2 (right) boundaries: z = {z_min_2:.3f} to {z_max_2:.3f} Å") | ||
|
|
||
| # Data from context: macroscopic average potential profile | ||
| CHECKPOINT_FILE = "./.mat3ra/checkpoint" | ||
| with open(CHECKPOINT_FILE, "r") as f: | ||
| checkpoint_data = json.load(f) | ||
| profile_interface = SimpleNamespace( | ||
| **checkpoint_data["scope"]["local"]["average-electrostatic-potential"]["average_potential_profile"] | ||
| ) | ||
| profile_left = SimpleNamespace( | ||
| **checkpoint_data["scope"]["local"]["average-electrostatic-potential-left"]["average_potential_profile"] | ||
| ) | ||
| profile_right = SimpleNamespace( | ||
| **checkpoint_data["scope"]["local"]["average-electrostatic-potential-right"]["average_potential_profile"] | ||
| ) | ||
|
|
||
| # Interface ESP profile | ||
| X = np.array(profile_interface.xDataArray) # z-coordinates (angstrom) | ||
| Y = np.array(profile_interface.yDataSeries[1]) # Macroscopic average V̄(z) | ||
|
|
||
| # Bulk material ESP profiles | ||
| X_left = np.array(profile_left.xDataArray) | ||
| Y_left = np.array(profile_left.yDataSeries[1]) | ||
| X_right = np.array(profile_right.xDataArray) | ||
| Y_right = np.array(profile_right.yDataSeries[1]) | ||
|
|
||
| def get_region_indices(x_data, x_min, x_max): | ||
| """Get array indices corresponding to coordinate range.""" | ||
| mask = (x_data >= x_min) & (x_data <= x_max) | ||
| indices = np.where(mask)[0] | ||
| if len(indices) == 0: | ||
| return 0, len(x_data) | ||
| return indices[0], indices[-1] + 1 | ||
|
Comment on lines
+101
to
+107
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Don’t fit the entire profile when a region is empty. Returning 🐛 Proposed fix (fail fast on empty regions) def get_region_indices(x_data, x_min, x_max):
"""Get array indices corresponding to coordinate range."""
mask = (x_data >= x_min) & (x_data <= x_max)
indices = np.where(mask)[0]
if len(indices) == 0:
- return 0, len(x_data)
+ return 0, 0
return indices[0], indices[-1] + 1
# Get indices for each slab region
slab1_start, slab1_end = get_region_indices(X, z_min_1, z_max_1)
slab2_start, slab2_end = get_region_indices(X, z_min_2, z_max_2)
+
+if slab1_end <= slab1_start:
+ raise ValueError(f"No points found in slab 1 region ({z_min_1}, {z_max_1}).")
+if slab2_end <= slab2_start:
+ raise ValueError(f"No points found in slab 2 region ({z_min_2}, {z_max_2}).")Also applies to: 94-100 🤖 Prompt for AI Agents |
||
|
|
||
| def fit_and_average(x_data, y_data, start_idx, end_idx): | ||
| """ | ||
| Fit linear regression to region and return average value, slope, and intercept. | ||
|
|
||
| The average of the fitted line equals the mean of y-values, | ||
| but fitting helps smooth out oscillations and validates linearity. | ||
| """ | ||
| x_region = x_data[start_idx:end_idx] | ||
| y_region = y_data[start_idx:end_idx] | ||
|
|
||
| if len(x_region) < 2: | ||
| avg = float(np.mean(y_region)) if len(y_region) > 0 else 0.0 | ||
| return avg, 0.0, avg | ||
|
Comment on lines
+109
to
+121
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Avoid silent 0.0 when the fit region is empty. Returning 🐛 Proposed fix (guard empty region) def fit_and_average(x_data, y_data, start_idx, end_idx):
@@
x_region = x_data[start_idx:end_idx]
y_region = y_data[start_idx:end_idx]
- if len(x_region) < 2:
- avg = float(np.mean(y_region)) if len(y_region) > 0 else 0.0
- return avg, 0.0, avg
+ if len(x_region) == 0:
+ raise ValueError(f"empty fit region for indices [{start_idx}, {end_idx})")
+ if len(x_region) == 1:
+ avg = float(y_region[0])
+ return avg, 0.0, avg🤖 Prompt for AI Agents |
||
|
|
||
| slope, intercept, r_value, _, _ = linregress(x_region, y_region) | ||
|
|
||
| # Average value of linear fit over the region | ||
| # V_avg = (1/L) * integral(slope*x + intercept) = slope*x_mid + intercept | ||
| x_mid = (x_region[0] + x_region[-1]) / 2.0 | ||
| avg_value = slope * x_mid + intercept | ||
|
|
||
| return float(avg_value), float(slope), float(intercept) | ||
|
|
||
| # Get indices for each slab region in interface profile | ||
| slab1_start, slab1_end = get_region_indices(X, z_min_1, z_max_1) | ||
| slab2_start, slab2_end = get_region_indices(X, z_min_2, z_max_2) | ||
|
|
||
| # Fit interface regions to get average ESP at interface | ||
| Va_interface, slope_a, intercept_a = fit_and_average(X, Y, slab1_start, slab1_end) | ||
| Vb_interface, slope_b, intercept_b = fit_and_average(X, Y, slab2_start, slab2_end) | ||
|
|
||
| # Get indices for slab regions in bulk profiles | ||
| slab1_start_left, slab1_end_left = get_region_indices(X_left, z_min_1, z_max_1) | ||
| slab2_start_right, slab2_end_right = get_region_indices(X_right, z_min_2, z_max_2) | ||
|
|
||
| # Fit bulk material profiles over the same z-ranges as their slabs | ||
| Va_bulk_left, _, _ = fit_and_average(X_left, Y_left, slab1_start_left, slab1_end_left) | ||
| Vb_bulk_right, _, _ = fit_and_average(X_right, Y_right, slab2_start_right, slab2_end_right) | ||
|
|
||
| # Calculate VBO using interface and bulk ESP values | ||
| # VBO = (interface difference) - (bulk difference) | ||
| VBO = (Vb_interface - Va_interface) - (Vb_bulk_right - Va_bulk_left) | ||
|
|
||
| print(f"Interface ESP Slab 1 (Va_interface): {Va_interface:.3f} eV") | ||
| print(f"Interface ESP Slab 2 (Vb_interface): {Vb_interface:.3f} eV") | ||
| print(f"Bulk ESP Left (Va_bulk): {Va_bulk_left:.3f} eV") | ||
| print(f"Bulk ESP Right (Vb_bulk): {Vb_bulk_right:.3f} eV") | ||
| print(f"Interface ∆V: {Vb_interface - Va_interface:.3f} eV") | ||
| print(f"Bulk ∆V: {Vb_bulk_right - Va_bulk_left:.3f} eV") | ||
| print(f"Valence Band Offset (VBO): {VBO:.3f} eV") | ||
|
|
||
| # Generate visualization plot | ||
| plt.figure(figsize=(10, 6)) | ||
| plt.plot(X, Y, label="Macroscopic Average Potential", linewidth=2) | ||
|
|
||
| # Highlight fitting regions | ||
| plt.axvspan(z_min_1, z_max_1, color="red", alpha=0.2, label="Slab 1 Region") | ||
| plt.axvspan(z_min_2, z_max_2, color="blue", alpha=0.2, label="Slab 2 Region") | ||
|
|
||
| # Plot fitted lines | ||
| if slab1_end > slab1_start: | ||
| x_fit1 = X[slab1_start:slab1_end] | ||
| y_fit1 = slope_a * x_fit1 + intercept_a | ||
| plt.plot(x_fit1, y_fit1, color="darkred", linestyle="--", linewidth=2, label="Fit Slab 1") | ||
|
|
||
| if slab2_end > slab2_start: | ||
| x_fit2 = X[slab2_start:slab2_end] | ||
| y_fit2 = slope_b * x_fit2 + intercept_b | ||
| plt.plot(x_fit2, y_fit2, color="darkblue", linestyle="--", linewidth=2, label="Fit Slab 2") | ||
|
|
||
| # Plot average ESP values | ||
| plt.axhline(Va_interface, color="red", linestyle=":", linewidth=2, label=f"Avg ESP Slab 1 = {Va_interface:.3f} eV") | ||
| plt.axhline(Vb_interface, color="blue", linestyle=":", linewidth=2, label=f"Avg ESP Slab 2 = {Vb_interface:.3f} eV") | ||
|
|
||
| plt.xlabel("z-coordinate (Å)", fontsize=12) | ||
| plt.ylabel("Macroscopic Average Potential (eV)", fontsize=12) | ||
| plt.title(f"Polar Interface VBO = {VBO:.3f} eV", fontsize=14, fontweight="bold") | ||
| plt.legend(loc="best", fontsize=10) | ||
| plt.grid(True, alpha=0.3) | ||
| plt.tight_layout() | ||
| # plt.show() | ||
|
|
||
| filename = f"polar_vbo_fit_interface.png" | ||
| plt.savefig(filename, dpi=150, bbox_inches="tight") | ||
| plt.close() | ||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Missing bounds check could cause IndexError.
If
material_1has the same number of atoms as the interface material (i.e.,n_left == len(z_elements)), accessingz_elements[n_left]on line 70 will raise anIndexError. Consider adding a guard.🛡️ Proposed fix
🤖 Prompt for AI Agents