core light-curve pipeline

This repository packages the photometric light-curve workflow that previously lived in the Almostly_final (2).ipynb notebook. The code is now structured as the importable package nups_core, exposing a configurable LightCurvePipeline class and a set of dataclasses that describe the processing parameters.

Installation

The pipeline depends on the scientific Python stack used in the original notebook:

• numpy
• pandas
• astropy
• photutils
• astroalign
• matplotlib

Create a virtual environment and install the requirements with pip:

python -m venv .venv
source .venv/bin/activate
pip install numpy pandas astropy photutils astroalign matplotlib

Running the pipeline

Edit Almostly_final.py (or create your own runner) to point to the desired dataset. Then execute:

python Almostly_final.py

The script instantiates LightCurvePipeline with the supplied configuration, executes the end-to-end calibration → alignment → photometry → ensemble detrending workflow, and writes the outputs to the configured output_dir.

Programmatic use

from nups_core import (
    PathsConfig,
    CalibrationConfig,
    AlignmentConfig,
    PhotometryConfig,
    EnsembleConfig,
    OutputConfig,
    PipelineConfig,
    LightCurvePipeline,
)

config = PipelineConfig(
    paths=PathsConfig(
        root_dir="/path/to/observation",
        output_dir="./pipeline_output",
    ),
)

result = LightCurvePipeline(config).run()
print(result.flux_matrix.shape, result.detrended_flux.shape)

The PipelineResult object captures the flux matrices, detrended series, star positions, covariates, and any artefacts saved to disk.

Outputs

The pipeline mirrors the notebook artefacts:

• Calibrated and (optionally) aligned FITS frames
• Detection preview image with apertures/annuli overlays
• Raw flux matrix CSV + time stamps
• Detrended relative-flux CSV
• Per-star detrended plots

Adjust the OutputConfig flags to enable or disable specific artefacts.