PySpecLines

Some useful scripts to compute integrated line fluxes and equivalent widths from galaxy spectra using 3 different methods:

• numerical (trapezoidal) integration
• Gaussian fit using a "standard" Levenberg-Marquardt algorithm
• Gaussian fit using an MCMC algorithm

Installing the package

To install PySpecLines you can use pip, which will take care of installing the required dependencies as well

pip install pyspeclines

To upgrade to the latest available version you can run

pip install pyspeclines --upgrade

Data format

FITS binary table

PySpecLines works with FITS tables. Each spectrum should be contained in a separate FITS file, which must contain the following columns:

• wl, wavelength array in ang
• flux, flux array in erg s^-1 cm^-2 ang^-1
• err, error array in erg s^-1 cm^-2 ang^-1

FITS header

PySpecLines also uses the following (optional) FITS header keywords:

• REDSHIFT, used to de-redshift a spectrum from the observed-frame to the rest-frame
• RA, right ascension of the object in deg, in the ICRS frame, used to correct for Galactic absorption
• DEC, declination of the object in deg, in the ICRS frame, used to correct for Galactic absorption

When both keywords RA and DEC are present, PySpecLines can de-redden the spectrum (passing the option --deredden) using the Galactic dust map of Schlegel, Finkbeiner & Davis (1998) recalibrated by Schlafly & Finkbeiner (2011), and the R_V=3.1 extinction curve of Fitzpatrick (1999).

JSON configuration file

A JSON file allows you to select and configure the emission lines to be measured. Some example JSON files are provided in the PySpecLines/files folder. The JSON file contains a dictionary of key : values where key labels the line (or group of lines) and values contains multiple entries.

Below we report some simple examples:

• single line:

      {
      "HeII4686" : {
        "wl_central": [4686.0], 
        "wl_range":[4680.0, 4681.0], 
        "continuum_left":[4672.0, 4679.0],
        "continuum_right":[4692.0, 4700.0]
      }
    }

  • wl_central is used as starting point for the line center in the Gaussian fitting
  • wl_range is the range of numerical integration of the line
  • continuum_left is the range of numerical integration of the left-continuum
  • continuum_right is the range of numerical integration of the right-continuum
  • when using Gaussian fitting, the range over which the continuum is fitted is [continuum_left[0], continuum_right[1]]

• line doublet:

    {
    "SII6716_SII6731" : {
        "wl_central": [6716.0, 6731.0], 
        "wl_range":[6710.0, 6723.0], 
        "exclude":[6710.0, 6721.0, 6726.0, 6740.0],
        "continuum_left":[6695.0, 6705.0],
        "continuum_right":[6740.0, 6750.0]
      }
    }

  • wrt to the example above, the key is composed of two labels separated by an underscore _
  • exclude allows to define regions ([exclude[0], exclude[1]], [exclude[2], exclude[3], ..., [exclude[2*i], exclude[2*i+1]]) excluded from the continuum fitting

• multiple kinematic components

    {
    "OIII5007N_OIII5007B" : {
        "wl_central": [5007.0, 5007.0], 
        "width": [100.0, 400.0], 
        "wl_range":[5000.0, 5014.0], 
        "exclude":[4995.0, 5020.0],
        "continuum_left":[4990.0, 5000.0],
        "continuum_right":[5020.0, 5034.0]
      }
    }

  • width allows to define multiple kinematic components, in this case a "narrow" (labelled OIII5007N) and a "broad" (labelled OIII5007B) component, whose starting widths must be set to different values.

• multiple lines with multiple kinematic components

    {
    "NII6548_HalphaN_HalphaB_NII6584" : {
        "wl_central": [6548.05, 6563.0, 6563.0, 6584.0], 
        "width": [100.0, 100.0, 400.0, 100.0], 
        "exclude":[6542.0, 6580.0],
        "wl_range":[6541.0, 6575.0], 
        "continuum_left":[6515.0,6542.0],
        "continuum_right":[6595.0,6610.0]
      }
    }

  • in this case we want to use the same width for different lines (NII6548, HalphaN and NII6584) and a different width for HalphaB. We thus use the same width value for NII6548, HalphaN and NII6584, as this will "tie" together their widths during the Gaussian fitting, while the width of HalphaB will be kept separate.

Examples

You can see the different available options running

pyspeclines --help

If the spectrum is provided in the observed frame, then you must provide a REDSHIFT keyword in the FITS header containing the object redshift.

• Compute the fluxes and EWs using numerical integration

  pyspeclines --file my_spectrum.fits --json-file  emission_lines_EWs_config.json
  

• Compute the fluxes and EWs using Gaussian fit (Levenberg-Marquardt)

  pyspeclines --file my_spectrum.fits --json-file  emission_lines_EWs_config.json --gaussian-fit
  

• Compute the fluxes and EWs using Gaussian fit (MCMC)

  pyspeclines --file my_spectrum.fits --json-file  emission_lines_EWs_config.json --gaussian-fit --use-PyMC --MCMC-samples 5000