Poraquê

Poraquê learns maps between the three-dimensional scalar fields of density-functional theory. Given only a crystal geometry it predicts the valence charge density and the kinetic energy density — no wavefunctions, no self-consistency cycle.

{POSCAR, INCAR, POTCAR} --analytic--> EXTCAR --Model 1--> CHGCAR --Model 2--> TAUCAR
                                                                                 |
                                                                        integrate v
                                                                              energy

The first step is closed-form; only the two field-to-field maps are learned. They are not unrelated regressions: the first is the Hohenberg–Kohn map, whose existence is a theorem, and the second is the kinetic energy density functional, the missing ingredient of orbital-free DFT.

Install

git clone https://github.com/seixas-research/poraque.git
cd poraque
pip install -e .

Python 3.11 or newer. Installing registers three console commands — poraque-train, poraque-inference and poraque-committee — which run from any directory once the environment is active. Each is the main() of the script of the same name under scripts/, so python scripts/poraque_train.py is equivalent to poraque-train and needs nothing installed.

Use

# 1. train one ext2chg and one chg2tau model on all structures
poraque-train --write-config configs/train_config.yaml
poraque-train --config configs/train_config.yaml

# 2. measure generalisation
poraque-train --config configs/train_config.yaml --kfold --k-folds 5

# 3. predict a structure that has never been computed
poraque-inference new_structure/ --output predictions/new_structure

Every predicted field is written in CHGCAR format and opens in VESTA.

Or drive it from ASE:

from ase.build import bulk
from poraque.calculator import Poraque

atoms = bulk("Au", "fcc", a=4.08, cubic=True)
atoms.calc = Poraque("models/poraque_models.pfno", potcar="POTCAR")
atoms.get_potential_energy()
print(atoms.calc.components)     # T_s, E_ext, alpha Z, E_H, E_xc, Ewald

Forces and stress are not implemented, so this is single points, not relaxations.

What is in here

Design points

• The external potential is computed natively. Poraquê reconstructs it from the POTCAR tables on any standard VASP output, matching a reference potential to a relative 5×10⁻⁵. There is no option to import one: the training input must be exactly what inference produces.
• Grids may differ between materials. One model serves all of them: the operator's weights live in Fourier-mode space, and batches are bucketed by grid shape.
• Constraints are structural where possible. For chg2tau, τ = τ_vW[ρ] + softplus(·) makes the Hoffmann-Ostenhof bound hold by construction rather than by penalty.
• Resampling is spectral. Fourier truncation is the exact band-limited projection for a plane-wave field; interpolation would alias and shift the electron count.
• CUDA, Apple Metal and CPU, selected automatically.

Status

Twelve gold supercells — ten 27-atom cells and two 32-atom cells, spanning four grid shapes. 5-fold cross-validation, whole structures held out:

The learned kinetic functional beats the analytic orbital-free functionals by a wide margin on this system — Thomas-Fermi scores 1.348 and von Weizsäcker 0.738 on the same fields, so chg2tau is 30× and 17× better respectively.

Cell size dominates the error

The aggregate above hides the only interesting thing in it. Split by cell size:

Held out, a 32-atom cell is 2.2–2.5× harder than a 27-atom one. That is the first transfer measurement this project has: with only two examples of that cell size, holding one out leaves a single sibling, and the operator has to extrapolate to a grid shape it has barely seen.

Within a familiar cell size, more data helps monotonically — the 27-atom numbers are the best yet recorded:

The three rows differ in protocol as well as in data — the 12-structure run holds out 3 structures per fold against 1 for the 5-structure run, and uses early stopping, which the earlier runs predate. Read the trend, not the third decimal.

Still one element. These numbers measure interpolation between geometries of gold and now, weakly, extrapolation across cell size. They say nothing about transfer to other chemistry. Growing the dataset remains the main open item — see docs/notes/roadmap.md.

Energies are not there yet. The total energy is a sum of terms of order 10⁴ eV whose physically relevant variation is a fraction of an eV per atom — a relative ~10⁻⁴ — and a field-level error of 2×10⁻² cannot survive that cancellation. Across the twelve structures the true spread is 0.27 eV/atom and the error on predicted differences is 0.29 eV/atom, a ratio of 1.06 with correlation r ≈ −0.1. That is an improvement on the previous 3× ratio, but an error equal to the signal and no correlation still means the predicted energy ordering carries no information. The energy module itself is validated against exact Madelung constants and uniform-electron-gas limits; it is the fields that are not yet accurate enough. See docs/source/energy/index.md.

License

MIT. See LICENSE.