The software package we developed for period-metallicity-luminosity (PML) relation (of RR Lyrae pulsators) validation purposes consists of several independently working python scripts (as well as some necessary modules). It should easily be extendable to other PML relations/approaches (for example Period-Wesenheit relations). The main PML validation script works in python 2.7, whereas the query scripts have been tested/used in 3.6. Future work includes updating the main script to work in 3.6 as well...
Additionally, we provide the user with a spectrum normalization and spectral analysis script, that allow the user to determine the fundamental parameters of their RR Lyrae pulsators (or other stars in general). These fundamental parameters can then be used for the PML validation.
The 'main_method_comp_script.py' script, referred to as the main PML validation/comparison script in following sections, makes use of several python modules to allow the user to assess the agreement between the GAIA and PML parallaxes. It has several functionalities to gain more insight on the sample you have selected, by allowing the user to print out several characteristics (e.g. the inferred distances to the stars, according to the PML relations). Moreover, the user should specify the options he/she wants, for the analysis of the agreement of the PML relations in the script itself (a parser interface has yet to be developed, although this might not be ideal, given the amount of options provided to the user).
The 'PML_relations.py' module contains the definitions of the PML relations whose agreement is tested. When used in the main script, it provides absolute magnitudes, given the required inputs for the PML relation. This should be refined when using other samples.
The 'Distance_Parallax.py' module contains the necessary definitions (distance modulus equation) to calculate the distances and parallaxes for the PML relations, taking into account interstellar attenuation. When used in the main script, it provides the PML parallaxes (and can provide explicit distances).
The ‘Tukey_Bland_Altman_Krouwer.py’ module (when used in the main script) generates the Tukey mean difference/Bland-Altman (BA) plots (Altman & Bland 1983; Bland& Altman 1986), as well as the Krouwer plots (Krouwer 2008). In the former the parallax differences are parametrized in function of the mean of the GAIA (DR2) parallax (Gaia Collaboration et al. 2016, 2018) and the PML parallax. In the latter they are parametrized in function of the GAIA (DR2) parallax. On top of that the module generates different plots that contain different inferences of possible biases, by means of several linear regression techniques (as well as their regression diagnostics), and provide one with the results (printed to a file) of the different statistical tests used to verify the assumption that the differences are normally distributed (a necessary assumption when making use of BA/Krouwer plots). For more information the user is deferred to the publications mentioned above.
The ‘Passing_Bablok.py’ module provides our implementation of the Passing-Bablok regression procedure (in python) used for PML validation (Passing & Bablok 2009). When used in the main script, it generates the different plots needed to analyze agreement, prints the necessary information for the user to a file, and tests the hypotheses β = 1 & α = 0 (i.e. the calculated PML relation parallaxes are similar to GAIA DR2 parallaxes). An important toint to be raised is that the inferred results rely on a (robust) linearity test: the cusum test as defined by Passing & Bablok (2009). In order to make any inferences about the agreement using the β,α parameters, the PML relation needs to pass this cusum test. If both (null) hypotheses are valid (i.e. the inferred parameters within the confidence interval are equal to 1 & 0), we can conclude the method is in agreement. Otherwise, it is not: deviations for β indicate proportional biases, deviations for α indicate offsets. For more information the user is deferred to the publication mentioned above.
Several query scripts based on the Astroquery python package (Ginsburg et al. 2018) were developed in order to easily extract data from the VizieR database (Ochsenbein et al. 2000) that will be used in the PML relation validation.
The create_data_Blahzko_csv.py script reads in a csv file of your liking that should contain the names of the stars of which you want to extract the Blazhko variability (according to the GCVS database, Samus et al. 2009). It then constructs a dat file (‘BLAZHKO_csv.dat’) that contains this Blazhko variability as well as the star names, that is read inby the main PML validation/comparison script.
The create_data.py script extracts the apparent magnitudes in the 2MASS Ks passband, as well as the AllWise W1 passband for a specific sample (our sample) of stars, defined by their names in the file itself. Moreover, it simultaneously extracts the GAIA DR2 parallaxes. It then constructs a dat file (‘W_K_plx.dat’) that contains all this information, and is read in by the main PML validation/comparison script.
The create_GAIA_data_csv.py script is specifically designed to easily extract GAIA DR2 parallaxes for the Dambis et al. (2013) sample of stars (although it can readily be extended to include query for parallaxes for other datasets), where it reads in two csv-files: one containing the cross-matched stars with the GAIA database and the other containing the full sample. It saves the parallaxes in a file called ‘GAIA_DATA_csv.dat’ that contains the parallaxes with corresponding uncertainties as well as the star names, that is read in by the main PML validation/comparison script.
Two different scripts are provided that allow the user to generate the necessary interstellar attenuation information for PML validation.
The 'Dusttable.py' script provides the user with a means to efficiently query the NASA/IPAC Galactic Dust Reddening and Extinction tool (https://irsa.ipac.caltech.edu/applications/DUST/), by creating a file called ‘sample.csv’ or ‘sample_CSV.csv’ depending on whether the user wants to obtain information on their manually defined sample (our predefined sample) or the csv-file containing their sample (e.g. the Dambis et al. (2013) sample that we used to test our software package). This file can then easily be uploaded to the tool, which subsequently provides you with the necessary attenuation information, which should be saved in a file called ‘Dustmap_output_CSV_table.txt’, in order to be read in by the main script (although this name can easily be changed).
The dereddening.py script takes photometric data obtained from VizieR in a votable format and calculates the interstellar attenuation/reddening with robust uncertainty estimates based on a monte carlo approach. Care has to be taken when selecting the actual data(set) used, as outliers might be present in the downloaded votable.
The 'specnorm_automized.py' script allows the user to normalize their spectra. It requires the spectra to be in the fits-file format.
The 'GSSP_fund_param_est.py' script provides the user with a estimate on the fundamental parameters of the targets, based on output from the Grid Search in Stellar Parameters (GSSP) software package (Tkachenko, 2015).