rdfadf 0.1.0

RDF/ADF computation tools with periodic-table-based labeling

rdfadf Package

1. Summary

rdfadf is for computing
RDF (Radial Distribution Function) and
ADF (Angular Distribution Function)
for crystalline and amorphous materials with high numerical accuracy.

This script was developed for educational purposes and has not been optimized for computational speed. As a result, it may become significantly slower when applied to very large systems with a high number of particles. This script is primarily built using the functionality provided by the Atomic Simulation Environment (ASE) https://ase-lib.org/.

Key features:

• Fully PBC-aware RDF / ADF computation
• Native support for ASE (ase.Atoms)
• Flexible selection of central and neighbor atoms using: element symbols / block (s,p,d,f) / group (G#) / period (P#) / All
• RDF:
  • Normalized per central atom
  • Optional 4πr² normalization
  • Cumulative counts → direct coordination number (CN)
• ADF:
  • Distance filtering (r_min, r_max)
  • Normalized per central atom
  • Optional Gaussian smoothing
• Optimized for research, MD trajectory analysis, and materials informatics workflows

---

2. Installation

Install the package from the project root directory:

pip install .

Uninstall:

pip uninstall rdfadf -y

---

3. Inputs

The package accepts ASE Atoms objects:

from ase.io import read

atoms = read("POSCAR")
atoms = read("structure.cif")
atoms = read("output.traj")

---

4. Label Specification

You can classify central and neighbor atoms using:

Category	Example	Meaning
Element	"Li", "O", "Co"	Specific element
All	"All"	All elements in the structure
Block	"s", "p", "d", "f"	s/p/d/f block
Group	"G16"	Group 16
Period	"P3"	Period 3

Examples:

• "Li"–"O" → Li–O RDF
• "Co"–"All" → Co around all elements
• "d"–"p" → d-block vs p-block atoms
• "G1"–"O" → Group 1 elements around O

---

5. RDF Function

compute_general_rdf(atoms, center_label, neighbor_label, ...)

Output includes:

r,
g_raw_norm,
g_smooth,
g_raw2,
g_smooth2,
cumulative_raw,
cumulative_smooth

cumulative_raw is the coordination number (CN) curve.

---

6. ADF Function

compute_general_adf(atoms, center_label, neigh1_label, neigh2_label, ...)

Supports:

• Distance filtering (r_min, r_max)
• Normalization per central atom
• Gaussian smoothing

---

7. Example (RDF)

from rdfadf import compute_general_rdf
from ase.io import read

atoms = read("POSCAR")

r, g_raw, g_smooth, g_raw2, g_smooth2, cn_raw, cn_smooth =     compute_general_rdf(
        atoms,
        center_label="Co",
        neighbor_label="O",
        rcut=6.0,
        bins=200,
        sigma=2.0
    )

---

8. Example (ADF)

from rdfadf import compute_general_adf
from ase.io import read

atoms = read("POSCAR")

theta, adf_raw, adf_smooth = compute_general_adf(
    atoms,
    center_label="Co",
    neigh1_label="O",
    neigh2_label="O",
    rcut=4.0,
    r_min=1.0,
    r_max=3.0,
    bins=180,
    sigma=2.0
)

---

9. Notes

• ADF computation may be more expensive than RDF (O(N × n_neighbors²))
• atoms.pbc = True is required for proper PBC handling
• Large MD trajectories require frame-by-frame analysis
• Adjust bins and rcut depending on the system

---

10. License

MIT License is recommended.