StepByStep.md

Step-By-Step

This file works as a Table of contents (TOC) to follow the procedures done to reproduce the information presented in the paper. The numbers on this TOC are the same as in the Jupyter notebooks. Each subseccion related to a Data, Additional or Figure file is indicated with a '=>'. It is important to keep in mind that one of the steps is not included in this repository (Step 5, where we run the CIGALE modelling in the Peregrine cluster). The information created at that step is located in a complementary Zenodo repositoty, where we add a script to generate the SEDs for a given galaxy inside the Seyfert Sample.

TOC

1. Create the Database Sample

• 1.1 Obtain a list of Seyfert galaxies from SIMBAD TAP => SMB_Dec_03_2020.csv
• 1.2 Use astropy (or TOPCAT) to cross-match SIMBAD galaxies and VCV galaxies => SMB_VCV_Dec_03_2020.csv
• 1.3 Obtain the bibcodes for the otype classifications in SIMBAD => BibcodesOtypes_Dec_03_2020.csv

2. Clean the Seyfert Sample

• 2.1 Remove different redshifts FIGURE 1
• 2.2 Separate the origin of the bibcode
• 2.3 Dealing with UNK bibcodes and Seyfert Types => NLS1_reclass.txt
• 2.4 Create the final sample of galaxies in terms of redshift and bibcodes => VCV_SMB_otypes.txt
• 2.5 Calculate the final numbers from the otypes TABLE 3

3. Obtain the SEDs for Database Sample (with a simplified Python script)

• 3.1 Get SEDs from CDS and NED with their respective bibcodes Bibcodes_SED.csv and eliminate duplicate measurements
• 3.2 Select photometric filters/bands
• 3.3 Save => SEDs for each galaxy

4. Create CIGALE Photometry file for Database Sample

• 4.1 Create equivalencies between NED and CDS tables
• 4.2 Transform photometry to CIGALE photometry file
• 4.3 Clean photometry points
• 4.4 Remove galaxies taking into account energy balance, at least 2 IR bands and 5 Opt bands => CIGPhot_BadEnergyBalance.tbl
• 4.5 Separating galaxies at different redshifts: Splited in ten groups of redshifts
• 4.6 Create CIGALE photometry files => CIGPhot_EnergyBal_NN.tbl and CIGPhot_EnergyBal_All.tbl

5. Run CIGALE with the parameters in TABLE 2 => CIGALEOutputs

• 5.1 Run Fritz setup
• 5.2 Run SKIRTOR setup
• 5.3 Run Fritz and SKIRTOR setup but only with two angles 30 and 70.
• 5.4 Run the No-AGN setup (fAGN=0).

6. Calculate the initial numbers of the sample used

• 6.1 Numbers from photometry TABLE 1
• 6.2 Numbers from VCV TABLE 3
• 6.3 Numbers from SMB TABLE 3
• 6.4 Numbers in both SMB and VCV, part of the values for TABLE 3

7. Join and clean CIGALE results

• 7.1 Join the ten redshift groups runs in one table
• 7.2 Reduced chi-square distributions FIGURE 3
• 7.3 Cleaning and save files => Cleanresults_*.fits
• 7.4 Calculate the counts for galaxy type TABLE 3

8. Example SEDs for SKIRTOR and Fritz

• 8.1 Select best and worst SED fittings => *best_model.fits
• 8.2 Plot the 5 SEDs for the best galaxy FIGURE 2 and the worst galaxy in the notebook

9. Verification with other estimations

• 9.1 Obtain the data from Vika et al.(2017)
• 9.2 Compare the parameters FIGURE 4

10. Pre-Analyse of data

• 10.1 Join clean results
• 10.2 Experiments with the parameters of the classifiers, assuming both classifications
• 10.3 Use GridSearchCV to find parameters for the classifiers

11. Analyse the results with Tables

• 11.1 Join clean results
• 11.2 Feature selection using RFECV FIGURE 5
• 11.3 Correlation-scores in the classification task with viewing angle and AGN disc luminosity TABLE 4
• 11.4 Correlation-scores in the classification task with ML TABLE 4
• 11.5 Predict classifications for S and SyG type galaxies TABLE 5

12. Analyse the results with Plots

• 12.1 Join clean results and transform to be used in the plots
• 12.2 Compare physical parameters from Fritz and SKIRTOR setups FIGURE 6
• 12.3 Compare difference in physical parameters with Type. We use both SMB and VCV classifications so Type-1=Sy1 and Type-2=Sy2 FIGURE 7. Additionally we compare the classifications also with the 30/70 setup FIGURE ALT 7
• 12.4 Compare other VCV classifications as Sy1.0, Sy1.2, Sy1.5, Sy1.8, Sy1.9 FIGURE 8 with SKIRTOR setup
• 12.5 Redshift evolution of physical parameters FIGURE 9
• 12.6 Accretion power (intrinsic luminosity of the disc AGN luminosity) for the different setups FIGURE 10

AppA. Verification of the narrow-line Seyfert 1 galaxies

• AppA.1 Join mock results
• AppA.2 Comparing AGN parameters FIGURE A1 AppB. Mock analysis of the output parameters
• AppB.1 Join information
• AppB.2 Check the reduced chi-square distributions
• AppB.3 Comparison between estimated and mock values for the SKIRTOR setup FIGURE B1 and Fritz setup FIGURE ALT B1