Planetary Orbital Spacing and Semi-Major Axes Statistical Dataset
Description
The dataset comprises planetary system alignment metrics derived from both observed planetary positions and predictions made by the GROSAU model. The GROSAU model assigns planets to discrete zones and bands based on semi-major axis spacing rules, allowing for a systematic comparison between predicted and observed orbital distances. The dataset integrates information at the planet-level, system-level, and band-level, providing a comprehensive framework for assessing model accuracy. At the planet level, each record includes the planet’s observed semi-major axis (AU), predicted AU according to the GROSAU model, residuals (predicted minus observed), and absolute residuals. Alignment within specific margins (±0.405, ±0.809, and ±1.618 AU) is indicated by Boolean flags, allowing for evaluation of strict and loose alignment criteria. System-level metrics are computed by aggregating planet-level data within each system. These include the total number of planets, the fraction of planets aligned within each margin, and the mean and median residuals. In addition, correlation and error statistics such as Pearson r, Spearman ρ, coefficient of determination (R²), root mean squared error (RMSE), and mean absolute error (MAE) are calculated to quantify the agreement between predicted and observed planetary positions. At the band level, planets are assigned to predicted zones and bands, and their alignment with the model is further assessed. Residuals and absolute differences are calculated for each planet within its respective band, enabling analysis of both local and global deviations from the predicted architecture. This dataset provides a robust framework for evaluating planetary system structure, testing the GROSAU model, and quantifying deviations between observed and predicted configurations. It is suitable for statistical analysis, visualization, and further modeling studies aimed at understanding the organization of planetary systems.
Files
Steps to reproduce
To reproduce the dataset and analyses, observed planetary positions are first combined with GROSAU predictions. Residuals and alignment flags are computed for the defined AU margins. System-level summaries are then generated by aggregating planet-level data, including alignment fractions and residual statistics. The following summary of steps can be taken using Python software statistical packages. (1) Load the Raw Data (pandas) Use the pandas library to import planet-level and system-level CSV files containing observed semi-major axes and GROSAU predictions. This step reads the datasets into memory and prepares them for analysis. (2) Assign Planets to Zones and Bands (custom functions in Python) Apply GROSAU-defined rules to map each planet to its predicted Zone and band. This hierarchical classification enables analysis of alignment patterns. See the associated publication for formula details. (3) Compute Residuals and Alignment Flags (pandas, numpy) Calculate the residuals as the difference between predicted and observed semi-major axes. Generate Boolean alignment flags indicating whether a planet falls within specified AU margins (±0.405, ±0.809, ±1.618 AU). These flags quantify alignment relative to the model. (4) Aggregate System-Level Metrics (pandas) Group planet-level data by system to calculate the number of planets, fraction of aligned planets, and mean and median residuals. This aggregation provides a summary of alignment and error for each planetary system. (5) Compute Correlation and Error Metrics (scipy and scikit-learn) Use scipy.stats and sklearn.metrics to calculate Pearson r, Spearman ρ, coefficient of determination (R²), root mean squared error (RMSE), and mean absolute error (MAE). These metrics quantify the agreement between predicted and observed planetary positions. (6) Generate Summary Tables and Visualizations (pandas, matplotlib, seaborn) Create tables summarizing alignment and residuals, and generate figures illustrating distributions, correlations, and band-level deviations. Visualization supports the interpretation and publication of results. (7) Export for Reproducibility (pandas) Save processed datasets, including residuals, alignment flags, and system-level summaries, to CSV files. This ensures that analyses can be independently reproduced and extended.
Institutions
- Boston University