SPC output

from akaitools import parse_spc

spc = parse_spc("calculation.spc")

The *_up.spc and *_dn.spc data files are located automatically next to the log file (using input_params.file as the stem). Override either path explicitly when needed:

spc = parse_spc(
    "calculation.spc",
    data_up="/data/up.spc",
    data_down="/data/dn.spc",
)

k-mesh and energy range

bsf = spc.spectral_up
if bsf is not None:
    km = bsf.kmesh
    print(f"Energy range : {km.energy_min:.4f} – {km.energy_max:.4f} Ry")
    print(f"Energy points: {km.n_energy}")
    print(f"k-path labels: {km.high_symmetry_indices}")  # e.g. {1: '(0 0 0)', 80: '(0 1 0)', ...}

BSF intensity matrix

SpectralFunction.data is a NumPy array of shape (n_energy, n_kpoints). None when n_sym_points == 0 (no k-path was computed).

import numpy as np

if bsf is not None and bsf.data is not None:
    print(f"Shape: {bsf.data.shape}")  # (n_energy, n_kpoints)

    # Energy axis in eV (1 Ry = 13.6057 eV)
    energy_Ry = np.linspace(bsf.kmesh.energy_min, bsf.kmesh.energy_max, bsf.kmesh.n_energy)
    energy_eV = energy_Ry * 13.6057

    # Intensity at the Fermi level (energy closest to 0 Ry)
    ef_idx = int(np.argmin(np.abs(energy_Ry)))
    print(f"BSF row at E_F: {bsf.data[ef_idx, :]}")

DataFrame export

SPCResult.to_dataframe() returns a tidy long-form frame with one row per (energy, k, spin) point, concatenating both spin channels when present:

df = spc.to_dataframe()
# columns: energy_Ry  k  spin  intensity

# Intensity along k for the energy point closest to E_F, spin up
ef_slice = df.query("spin == 'up'")
ef_slice = ef_slice[ef_slice["energy_Ry"].sub(0.0).abs() == ef_slice["energy_Ry"].sub(0.0).abs().min()]

Channels with no data (None, or an uncomputed k-path) are omitted; to_dataframe() returns an empty frame with the same columns when neither channel has data.

Plotting the spectral function

from akaitools.plotting import plot_bsf

fig = plot_bsf(spc, spin="up", energy_unit="eV")
fig.savefig("bsf.png", dpi=150)

See Plotting for the full set of options (spin filtering, colormap, vmax, and the placeholder behavior when a channel's SpectralFunction.data is None).

SPC parameters

sp = spc.spc_params
print(f"Energy window : {sp.ew:.4f} Ry")
print(f"Broadening eta: {sp.eta:.4f} Ry")
print(f"k-points (BZ) : {sp.nk}")
print(f"Directions    : {sp.nd}")

for prop in spc.atomic_properties:
    print(f"  {prop.element}: m_spin={prop.spin_moment:.4f} μB")