# Agent input data 输入为 mossbauer_spin_crossover_input.zip,解压到 /app/input 后包含: 1. mossbauer_spectra.csv:18432 行(18 spectra×1024 channels);字段为 spectrum_id、temperature_K、replicate、acquisition_order、channel、monitor_counts、transmitted_counts、source、absorber_batch。计数为整数,无速度列和组分标签。 2. velocity_calibration_peaks.csv:108 行(18 spectra×6 α-Fe lines);含 observed_channel、certified_velocity_mm_s 及其 SD,用于逐谱通道—速度标定。 3. hyperfine_reference.csv:3 行;给出 LS-Fe(II)、HS-Fe(II)、Fe(III) 的氧化态/自旋归属、isomer shift、quadrupole splitting、FWHM 和独立来源说明,速度单位 mm/s。 4. recoilless_fraction_calibration.csv:18 行(6 temperatures×3 species);含 recoilless_fraction、SD、校准重复数和来源,分数无量纲。 5. analysis_config.json:温度序列、α-Fe 认证线、组分顺序、Fe(II) 高自旋分数口径和公开 QC 阈值。 6. README_input.md:实验口径、数据生成基础、计数/基线/标定变异和数据敏感性说明。CSV 为 UTF-8;温度 K、速度 mm/s。 ## Annotator input notes # Variable-temperature 57Fe Mössbauer input The package contains 18 transmission spectra measured at six temperatures with three independent acquisitions per temperature. Detector channels have not been converted to velocity. Each spectrum has a separate six-line alpha-iron calibration table. The sample can contain low-spin Fe(II), high-spin Fe(II), and a minor Fe(III) doublet. The hyperfine table comes from independent reference fits. Recoilless fractions were measured with matched-thickness reference absorbers and must be linked by temperature and species. Relative spectral area is therefore not, by itself, a mole fraction. Counts include Poisson variation, acquisition-to-acquisition baseline changes, small velocity-scale changes, and modest hyperfine variation. No row contains the true component population, transition temperature, thermodynamic parameters, acceptance result, or an artificial exclusion label. This is a physics-based synthetic benchmark dataset. It does not reproduce private experimental data. The line model, count statistics, calibration drift, absorber response, and thermodynamic population law follow conventional transmission Mössbauer practice; the exact realization is reproducible from the reference source retained by the benchmark owner. ## File inventory | file | bytes | |---|---:| | `README_input.md` | 1297 | | `analysis_config.json` | 816 | | `hyperfine_reference.csv` | 356 | | `mossbauer_spectra.csv` | 1113707 | | `recoilless_fraction_calibration.csv` | 1230 | | `velocity_calibration_peaks.csv` | 6552 | ## Computed numeric profile ### `hyperfine_reference.csv` Rows: 3; columns: 7. | column | min | max | missing | |---|---:|---:|---:| | `isomer_shift_mm_s` | 0.35 | 1.05 | 0 | | `quadrupole_splitting_mm_s` | 0.55 | 2.2 | 0 | | `fwhm_mm_s` | 0.28 | 0.3 | 0 | ### `mossbauer_spectra.csv` Rows: 18432; columns: 9. | column | min | max | missing | |---|---:|---:|---:| | `temperature_K` | 195 | 330 | 0 | | `replicate` | 1 | 3 | 0 | | `acquisition_order` | 1 | 18 | 0 | | `channel` | 0 | 1023 | 0 | | `monitor_counts` | 4842840 | 5156704 | 0 | | `transmitted_counts` | 4026271 | 4799271 | 0 | ### `recoilless_fraction_calibration.csv` Rows: 18; columns: 6. | column | min | max | missing | |---|---:|---:|---:| | `temperature_K` | 195 | 330 | 0 | | `recoilless_fraction` | 0.495 | 0.835 | 0 | | `recoilless_fraction_sd` | 0.008 | 0.01 | 0 | | `calibration_replicates` | 4 | 4 | 0 | ### `velocity_calibration_peaks.csv` Rows: 108; columns: 7. | column | min | max | missing | |---|---:|---:|---:| | `temperature_K` | 195 | 330 | 0 | | `replicate` | 1 | 3 | 0 | | `observed_channel` | 53.619829 | 970.26315 | 0 | | `certified_velocity_mm_s` | -5.33 | 5.33 | 0 | | `certified_velocity_sd_mm_s` | 0.004 | 0.004 | 0 |