Looking for Anomalies in Fink data ....
To get started, make sure to run the setup.py (python setup.py install) file to set up the necessary functions. You'll also need to download the fink_utils package.
Next, download the Fink data from: Fink Portal. You can see an example with the data we used in our work here: Example Data (data1).
After that, you can perform data transfer or data reduction and feature extraction using the file data-transfer_GP.py. This process may take about half an hour for the (data1) file. Please note that this file fills the gaps with Gaussian Process regression. We also have a second file, data-transfer.py, which uses upper limits and weighted mean to fill gaps.
(Note: These two methods are direct applications on all objects. You can find explanations for each step on one object in the notebooks, the .ipynb versions.)
Simultaneously, go to scripts/classification_functs/Run_jobs_th.py and run the Python file. This step may take a significant amount of time, but it only needs to be done once after you download your data. It saves the classification of objects in the Classifications_arch folder so you don't have to call the server every time you want to see the classification.
If you are working with (data1), you can find the Classifications_arch folder here: Google Drive link.
Now, we start the analysis. We have two different files: one with optimization (retrieve_data-exclud_matches.ipynb) and one without optimization (retrieve_data.ipynb).
Here is a link to the six-month internship on this package, where we explain all the details, algorithms, and some results.