Skip to main content

Fink SNAD Anomaly Detection Model

Project description

Fink anomaly detection model

A set of modules for training models for finding anomalies in photometric data. There are currently two entry points via the console: fink_ad_model_train and get_anomaly_reactions.

fink_ad_model_train

The module trains the AADForest model using expert reactions from the C055ZJJ6N2AE channels in Slack and -1001898265997 in Telegram. It creates the following files:

  • _g_means.csv and _r_means.csv -- averages over the training dataset;
  • _reactions_g.csv and _reactions_r.csv -- training datasets for additional training of the AADForest algorithm, based on expert reactions from Slack and Telegram channels;
  • forest_g_AAD.onnx -- model for _g filter
  • forest_r_AAD.onnx -- model for _r filter

optional arguments:

--dataset_dir DATASET_DIR Input dir for dataset (default: './lc_features_20210617_photometry_corrected.parquet')

--n_jobs N_JOBS
Number of threads (default: -1)

usage: fink_ad_model_train [-h] [--dataset_dir DATASET_DIR] [--n_jobs N_JOBS]

get_anomaly_reactions

Uploading anomaly reactions from messengers. It creates the following files:

  • _reactions_g.csv and _reactions_r.csv -- training datasets for additional training of the AADForest algorithm, based on expert reactions from Slack and Telegram channels;

optional arguments:

--slack_channel SLACK_CHANNEL Slack Channel ID (default: 'C055ZJJ6N2AE')

--tg_channel TG_CHANNEL Telegram Channel ID (default: -1001898265997)

usage: get_anomaly_reactions [-h] [--slack_channel SLACK_CHANNEL] [--tg_channel TG_CHANNEL]

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

fink_anomaly_detection_model-0.4.31.tar.gz (8.1 kB view details)

Uploaded Source

Built Distribution

File details

Details for the file fink_anomaly_detection_model-0.4.31.tar.gz.

File metadata

File hashes

Hashes for fink_anomaly_detection_model-0.4.31.tar.gz
Algorithm Hash digest
SHA256 c81fc9e3740f679c1154e8aa3760d83820dcae1b5802fb68a91e95688a175619
MD5 852b6b35357cc9bbcf6e722e207927bf
BLAKE2b-256 bd85d1137848cb53a1ef943562828b97cc6df2dcfdd3480aa488271a103d08da

See more details on using hashes here.

File details

Details for the file fink_anomaly_detection_model-0.4.31-py3-none-any.whl.

File metadata

File hashes

Hashes for fink_anomaly_detection_model-0.4.31-py3-none-any.whl
Algorithm Hash digest
SHA256 64d2aa63598f66ae711c12f992e9fc386fae898be3b463b781154860aad1e132
MD5 84e26c50a557ab7f9ed92e46b057781e
BLAKE2b-256 cbc452ff1e6a20731ea0d7e21498f069affa1905fa0a07cd3cf3898b707a2f7f

See more details on using hashes here.

Supported by

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