Skip to main content

Kalman Real-time Evidence Monitoring Extractor for TART

Project description

kremetart

Kalman Real-time Evidence Monitoring Extractor for TART.

Holoscan-driven applications to image — and, in time, detect transients in — data from the Transient Array Radio Telescope (TART). The streaming inference machinery is GPU-resident (CuPy + Holoscan); a CUDA GPU is highly recommended.

Installation

pip install kremetart          # lightweight: hip-cargo + typer only
pip install "kremetart[full]"  # full native stack (CuPy/Holoscan, healpy, xarray, tart-tools, …)

The lightweight install is enough to invoke any command: when the heavy native deps are absent, the CLI transparently dispatches the command into the project's container image. The full install runs everything natively.

Usage

kremetart --help

Producing a movie from the test data

The bundled TART snapshots under tests/data/ are gitignored and fetched on demand from Google Drive the first time you run the test suite. Populate them once (set KREMETART_OFFLINE=1 to skip the download):

uv run pytest tests/test_smoovie.py -q

Then render the snapshots into a HEALPix all-sky movie with smoovie (needs a CUDA GPU and ffmpeg on PATH). smoovie images one frame per sub-integration — the bundled data is nine ~1-minute snapshots of ~1-second sub-integrations, i.e. 540 frames — so render the whole sequence at a watchable frame rate:

uv run kremetart smoovie --hdf-dir tests/data --movie /tmp/tart.mp4 --nside 64 --fps 12 --correct-gains --overlay-catalog

Rendering dominates the runtime (~0.8 s/frame for the three movies), so the full 540-frame render takes a few minutes. For a quick look, cap the number of frames with --nframes — but note this caps the total imaged sub-integrations, and only a few seconds of sky barely moves, so use a generous value (e.g. the first ~2 minutes of sky):

uv run kremetart smoovie --hdf-dir tests/data --movie /tmp/preview.mp4 --nside 64 --fps 12 --nframes 120 --correct-gains --overlay-catalog

Either command writes the dirty all-sky movie (/tmp/tart.mp4) alongside *.filtered.mp4 (the IWP filtered flux), *.znorm.mp4 (the normalised innovation), and a durable *.zarr holding the dirty/filtered/znorm (TIME, PIX) maps. smoovie refuses to overwrite an existing <movie>.zarr — pass --overwrite to replace it. It works on any directory of TART *.hdf snapshots; see kremetart smoovie --help for all options.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

kremetart-0.0.1.tar.gz (41.3 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

kremetart-0.0.1-py3-none-any.whl (54.0 kB view details)

Uploaded Python 3

File details

Details for the file kremetart-0.0.1.tar.gz.

File metadata

  • Download URL: kremetart-0.0.1.tar.gz
  • Upload date:
  • Size: 41.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for kremetart-0.0.1.tar.gz
Algorithm Hash digest
SHA256 eee766b4a42493a8e46388a6af2932d40585b8e4b04d7f9e08c94481cf5e0d34
MD5 934031bc155b00dff3e3bdc68dd22c43
BLAKE2b-256 1600c94c9a14c73525e94af5b2c3e1db209fd3a1f7fb67bfa74e2287dd250698

See more details on using hashes here.

Provenance

The following attestation bundles were made for kremetart-0.0.1.tar.gz:

Publisher: publish.yml on landmanbester/kremetart

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file kremetart-0.0.1-py3-none-any.whl.

File metadata

  • Download URL: kremetart-0.0.1-py3-none-any.whl
  • Upload date:
  • Size: 54.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for kremetart-0.0.1-py3-none-any.whl
Algorithm Hash digest
SHA256 0720929aa4856a75db9efc64b217b3b27091ef64b8831b1fbb40e117263916df
MD5 a7f7d34430e27f62ceff1fbb7f78bb7e
BLAKE2b-256 372180a1e9e1e557d4ed52b2f7652412059ed5f22c1b88e26cc05de8042f4e2f

See more details on using hashes here.

Provenance

The following attestation bundles were made for kremetart-0.0.1-py3-none-any.whl:

Publisher: publish.yml on landmanbester/kremetart

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page