The source code of finding strong galaxy-galaxy lens from SDSS spectra (arxiv: 2202.08439)
Python
eBOSS_spetra_lens
The source code of finding strong galaxy-galaxy lens from SDSS spectra (arxiv: 2202.08439)
Total five-step of this project.
1. You need to run "Serching_Galaxy.py" to select out all possible useable galaxies spectra. Notice, you need to download the SDSS spectra "spPlate.fits" and "spZbest.fits" files to the "PATH" folder.
2. Run "SDSS_Spetra_Produce.py" to Bin the redshift for spectra generated in step.1, and cut the spectra in a uniform format, which same as the shape of CNN models. It will produce 4 pieces of data, corresponding train, validation, test, and prediction samples.
3. According to the output data of step.3, run the "Build_CNN_Data.py" to generate CNN's labeled positive and negative samples. The positive samples will be adding artificial emission lines.
4. Run the "model.py" to train the CNNs.
5. After training CNNs' modes, run "pred.py" to predict the value of PL, ZE, and ZG. Predict results will be saved in csv files.