Fink Science

This repository contains science modules used to generate added values to alert collected by the Fink broker.

ZTF alert stream

It currently contains:

Source	Field in Fink alerts	Type	Contents
fink_science/xmatch	`cdsxmatch`++	string	Counterpart (cross-match) from any CDS catalog or database using the CDS xmatch service. Contains also crossmatch to the General Catalog of Variable Stars and the International Variable Star Index, 3HSP, 4LAC DR3.
fink_science/random_forest_snia	`rf_snia_vs_nonia`	float	Probability to be a rising SNe Ia based on Random Forest classifier (1 is SN Ia). Based on https://arxiv.org/abs/2111.11438
fink_science/snn	`snn_snia_vs_nonia`	float	Probability to be a SNe Ia based on SuperNNova classifier (1 is SN Ia). Based on https://arxiv.org/abs/1901.06384
fink_science/snn	`snn_sn_vs_all`	float	Probability to be a SNe based on SuperNNova classifier (1 is SNe). Based on https://arxiv.org/abs/1901.06384
fink_science/microlensing	`mulens`	float	Probability score to be a microlensing event by LIA
fink_science/asteroids	`roid`	int	Determine if the alert is a Solar System object
fink_science/kilonova	`rf_kn_vs_nonkn`	float	probability of an alert to be a kilonova using a Random Forest Classifier (binary classification).
fink_science/nalerthist	`nalerthist`	int	Number of detections contained in each alert (current+history). Upper limits are not taken into account.
fink_science/ad_features	`lc_*`	dict[int, array]	Numerous light curve features used in astrophysics.
fink_science/agn	`rf_agn_vs_nonagn`	float	Probability to be an AGN based on Random Forest classifier (1 is AGN).

You will find README in each subfolder describing the module.

These modules are being tested for Rubin era on the LSST-DESC ELASTiCC data challenge:

Source	Field in Fink alerts	Type	Contents
fink_science/agn_elasticc	`rf_agn_vs_nonagn`	float	Probability to be an AGN based on Random Forest classifier (1 is AGN).
fink_science/random_forest_snia	`rf_snia_vs_nonia`	float	Probability to be a rising SNe Ia based on Random Forest classifier (1 is SN Ia). Based on https://arxiv.org/abs/2111.11438
fink_science/snn	`snn_snia_vs_nonia`	float	Probability to be a SNe Ia based on SuperNNova classifier (1 is SN Ia). Based on https://arxiv.org/abs/1901.06384
fink_science/snn	`broad`	array[float]	Broad classifier based on SNN. Returns [class, max(prob)].
fink_science/cats	`fine`	array[float]	Fine classifier based on the CBPF Algorithm for Transient Search. Returns [class, max(prob)].

Source	Field in Fink alerts	Type	Contents
fink_science/t2	`t2`	array[float]	Classifier based on Transformers. Based on https://arxiv.org/abs/2105.06178

Learn how to design your science module, and integrate it inside the Fink broker.

If you want to install the package (broker deployment), you can just pip it:

pip install fink_science