Notebook Atlas
Each project keeps its analysis in numbered notebooks. This page lists them in reading order and points to the documentation page that explains the same step. Notebook outputs are not reproduced here: the documentation quotes only values that are traceable to an exported result file.
AstroSpectro
Repository: PhD-Brown/AstroSpectro, folder notebooks/.
| Notebook | Purpose | Documentation |
|---|---|---|
00_master_pipeline.ipynb | Runs the whole pipeline from data selection to a session report | Quick Start · Master Pipeline |
01_download_spectra.ipynb | Interactive download of LAMOST DR5 spectra | Data download |
02_tools_and_visuals.ipynb | Interactive spectrum analyzer and visualization tools | Visualization |
03_scientific_validation.ipynb | Checks that the extracted features behave as spectroscopy predicts | Scientific Validation |
04_shap_interpretability.ipynb | SHAP analysis of the classifier | SHAP Interpretability |
dimred/phy3500_01_pca.ipynb | PCA of the PHY-3500 feature table (G1) | PCA |
dimred/phy3500_02_umap_tsne.ipynb | UMAP, t-SNE and HDBSCAN (G1) | UMAP & HDBSCAN · t-SNE |
dimred/phy3500_03_autoencoder.ipynb | Autoencoder (G1) | Autoencoder |
The three dimred/ notebooks belong to the PHY-3500 · G1 generation of data; the other notebooks work on the pipeline and the reference classifier. The two generations are described in Data & Datasets.
AstroVision
Repository: PhD-Brown/AstroVision, folder notebooks/.
| Notebook | Purpose | Documentation |
|---|---|---|
00_eda_galaxy10.ipynb | Exploratory analysis of Galaxy10 DECaLS | Galaxy10 DECaLS |
01_galaxy10_cnn.ipynb | Baseline CNN and EfficientNet training, with Grad-CAM | Baseline CNN · EfficientNet-B0 |
02_results_analysis.ipynb | Comparison of the first models, per-class results and errors | Model Benchmarking |
03_dinov2_features.ipynb | Zero-shot DINOv2 features and linear probe | DINOv2 — Linear Probe |
04_dinov2_finetune.ipynb | Fine-tuning DINOv2 | DINOv2 — Fine-tuned |
05_crossmatch_analysis.ipynb | SDSS DR16 cross-match and colour analysis | SDSS DR16 Crossmatch · Physical Interpretation |
06_segmentation.ipynb | U-Net segmentation on Otsu pseudo-labels | U-Net Segmentation |
07_synthesis.ipynb | Synthesis of morphology, photometry and segmentation | Physical Interpretation |
08_multimodal_fusion.ipynb | Late-fusion multimodal classifier and its ablation | Late-Fusion Multimodal |
09_uncertainty_calibration.ipynb | MC Dropout, temperature scaling, conformal prediction | Uncertainty Quantification |
10_attention_morphometry.ipynb | CAS/Gini/M20 measured inside DINOv2 attention masks | Attention & GradCAM · Non-parametric Morphometrics |
AstroVision scores exist only in notebook outputs, not in exported metric files; the AstroVision pages state which values are read from which notebook (see _facts/astrovision/FACTS_INDEX.md).
ξ Dark Energy
Repository: PhD-Brown/xi-dark-energy-bao-mcmc, folder notebooks/.
| Notebook | Purpose | Documentation |
|---|---|---|
00_project_overview.ipynb | Tour of the final results; does not rerun the MCMC | Key Findings |
01_data_inspection.ipynb | Checks that the local data files are present and inspects their dimensions | Data Overview |
02_results_review.ipynb | Loads the final result tables and summarises them | Parameter Constraints |
03_mcmc_diagnostics.ipynb | Convergence diagnostics from the saved chains (chains are not committed) | Convergence Diagnostics |
The MCMC itself is run by scripts (src/xi_protocol1_run.py, src/xi_protocol1b_bao_run.py), not by a notebook; see Quick Run.
See also: Repository Map · Methods Index.