Dataset: peripheral defocus across spectacle lens form, material, pupil and accommodation (GRIN ray-tracing eye)

Raw model-output CSVs accompanying the two-part manuscript on peripheral defocus and myopia, generated by the open grin-defocus-eye engine v1.0.0 (GRIN ray-tracing schematic eye; AVOCADO gradient-index crystalline lens; minimum-RMS / power-vector readout). Each row is a direct engine readout — no post-processing. Conventions. Dioptres (D); sign clinical (myopia < 0, hyperopia > 0); RPRE(θ) = M(θ) − M(0) (relative peripheral refractive error); J0 > 0 = with-the-rule astigmatism. Field angle = retinal eccentricity (deg). Peripheral readout validated against a homogeneous Le Grand oracle to ≤ 0.03 D out to ~20°; 30° reported with a caveat; 40° out of trust. License: CC–BY–4.0. Part 1 — bare-eye validation & biometry (no correction). How the uncorrected eye’s peripheral refraction depends on refraction and biometry (separate protocol: pupils 3/5/7/9 mm, angles to 40°): data/a0_baseline_rpre.csv — emmetrope bare-eye RPRE & astigmatism (J0/J45) vs eccentricity 0–30°. data/p1_ametropia_series.csv — bare-eye RPRE@20/40° per pupil across the ametropia series (axial H1 vs refractive H2). data/p1_biometry.csv — bare-eye RPRE distance/near per biometric-factor configuration. data/p1_A2_axial_length.csv, p1_A3_keratometry.csv, p1_A4_pupil.csv, p1_A5_acd.csv — one-factor RPRE sensitivities (axial length, corneal power, pupil, anterior chamber depth). data/p1_A6_AL_x_K_grid.csv, p1_alk_grid.csv — axial-length × keratometry grid. data/p1c_A7_accommodation_validation.csv, p1c_A7_accommodation_x_AL.csv, p1c_A7_decomposition.csv — accommodation: central power-gain check, RPRE@40° vs axial length, sub-term decomposition. data/p1d_grin_sensitivity.csv — RPRE@40° sensitivity to GRIN-lens conic perturbations. Part 2 — spectacle peripheral defocus. data/lensform_sweep.csv (720 rows): one row per power, lens form, material, accommodation state, pupil and eccentricity. The lens is designed at distance (central M(0) = 0); the near state solves accommodation through the lens (effectivity) so a 33 cm object is centrally focused; the periphery is read with the collimated, validated reader in both states. The bare-eye reference is included as form=none. Columns: SR — nominal spherical-equivalent the eye is built for (D): −6 / −3 / +3 / +6. eye_M0 — actual central refraction of the bare eye (D). form — none (bare-eye reference) / trial (symmetric sphere) / sphere (best-form) / asphere (front-aspheric point-focal). material, n_e — lens material (cr39 = 1.50, mid156 = 1.56, hi167 = 1.67) and its refractive index ne. cond — distance (A = 0) or near33 (accommodated for 33 cm through the lens). A_acc — ocular accommodation applied (D). Pf, conic, A4, A6, tc, front_R, back_R — designed lens parameters: front-surface power (D), conic Q, even-aspheric coefficients, centre thickness (mm), surface radii (mm). pupil (2 / 4 / 6 mm), angle (10 / 20 / 30° eccentricity). M0 — central refraction in this state (D); RPRE = M(θ) − M(0) (D); J0 = oblique-astigmatism power vector (D). The aspheric lens is point-focal, i.e. its J0 ≈ the form=none baseline (it adds ~no astigmatism), not zero. Provenance. Generated by grin-defocus-eye v1.0.0, deposited separately as software: https://doi.org/10.5281/zenodo.20719403 (the bare-eye engine is unchanged from the validated release). See DATA_README.md in the archive for the full data dictionary and reproduction commands. How to cite. Cite both the software (10.5281/zenodo.20719403) and this dataset (DOI assigned by Zenodo on upload).