scintkit

Simple tools for working with ScintPi and GNSS scintillation data.

This repository provides utilities and pipelines for:

• converting raw binary data to Parquet
• adding derived products (TEC, S4, sigma_phi, etc.)
• running processing workflows
• plotting ScintPi data

The Level 3 product schema and automatic quality-flag rules are documented in docs/lvl3.md.

src/scintkit/reading/         # binary readers and data loading
src/scintkit/preprocessing/   # formatting and preprocessing
src/scintkit/services/        # core computations (TEC, S4, phase detrending)
src/scintkit/utils/           # helper utilities
src/scintkit/pipelines/       # end-to-end processing pipelines
tests/                        # test scripts and example notebooks

Clone with git:

git clone https://github.com/qwsae10/scintkit.git
cd scintkit

Or download as a ZIP from GitHub and extract it.

Editable install for development:

python -m pip install -e .

This makes the package importable as scintkit.

You can import and run core processing functions directly.

from scintkit.pipelines.auto import process

process("example.bin.zip")

This will:

1. convert raw data to parquet
2. apply preprocessing
3. compute derived products
4. output lvl3 files

The maintained station registry can identify a receiver from coordinates:

from scintkit.data import identify_station

station = identify_station(latitude=32.9919, longitude=-96.7573)
print(station["Code"])  # US-TX1

It can also read legacy coordinate filenames or SC4 station prefixes:

station = identify_station(
    filename="scintpi3_20241011_1200_96.7573W_32.9919N_v326f_lvl0.pq"
)

The function returns the matching CSV row as a dictionary. It returns None when no station is within 3 km. Pass max_distance_km= to change that limit.

See:

examples/compare_oct11.ipynb

This notebook shows how to:

• load raw data
• run processing steps
• compare outputs