Defining filters

bayesn-td includes a selection of filters and standards for convenience. Alternatively, you can use your own through a simple YAML file, allowing new or updated filters to be easily implemented independently of updates to the package.

Built-in filters

bayesn-td includes the following filters, grouped by instrument, survey, or system. The name shown is the string to use when referring to the filter within bayesn-td (e.g. g_PS1 for Pan-STARRS 1 g-band). If the names in your data files do not match these, you can provide a band_map to match them up, as discussed in Fitting lensed SNe within Python.

  • Standard

    • Stritzinger UBVRI filters

      • Names: U, B, V, R, I

      • Source: Stritzinger+11, 2005PASP..117..810S

    • 2MASS Peters Automated Infrared Imaging Telescope JHK filters

      • Names: J, H, K

      • Source: Cohen+03, 2003AJ….126.1090C

    • Persson YJHK filters

      • Names: Y_P, J_P, H_P, K_P

      • Source: 1998AJ….116.2475P

  • CSP

  • DECam

  • HST

  • JWST

    • NIRCam broad-band filters

      • Names: F090W, F115W, F150W, F200W, F277W, F356W, F444W

      • Source: STScI

  • LSST

  • PS1

    • Pan-STARRS 1 grizy filters

      • Names: g_PS1, r_PS1, i_PS1, z_PS1, y_PS1

      • Source: Tonry+12, 2012ApJ…750…99T

  • Roman

    • Wide Field Instrument (WFI) filters

      • Names: roman::Z, roman::Y, roman::J, roman::H, roman::F, roman::R

  • SWIFT UVOT

    • UVOT UBV and UV filters

      • Names: U_SWIFT, B_SWIFT, V_SWIFT, UVW1, UVW2, UVM2

      • Source: Poole+08, 2008MNRAS.383..627P

  • USNO

    • United States Naval Observatory 40-inch telescope u’g’r’i’z’ filters

      • Names: u_prime, g_prime, r_prime, i_prime, z_prime

      • Source: Fukugita+96, 1996AJ….111.1748F; Smith+02, 2002AJ….123.2121S

  • ZTF

    • Zwicky Transient Facility gri filters

      • Names: p48g, p48r, p48i

      • Source: Bellm+19, 2019PASP..131a8002B

  • ANDICAM

    • ANDICAM at Cerro Tololo Inter-American Observatory YJHK filters

      • Names: Y_AND, J_AND, H_AND, K_AND

For a complete listing of all available filters, inspect the bundled filters.yaml file in the bayesn-td package directory.

Specifying custom filters

You can add custom filters beyond those built in by providing a filter YAML file via the filter_yaml argument when initialising the model, or the --filters command-line argument. Custom filters are included alongside the built-in ones, so you can mix and match. If you give a custom filter the same name as a built-in one, your custom filter will take precedence.

The filter YAML file should have the following structure:

standards_root: /path/to/standards
standards:
  vega:
    path: vega_standard.fits
  bd17:
    path: bd17_standard.fits
filters_root: /path/to/filters
filters:
  my_band_1:
    magsys: ab
    magzero: 0
    path: my_band_1_response.dat
    lam_unit: nm
  my_band_2:
    magsys: vega
    magzero: 0
    path: my_band_2_response.dat

  • standards_root — root directory for standard spectra (optional; paths can be absolute).

  • standards — defines reference spectra for magnitude systems. Each entry needs a path to a FITS or text file with WAVELENGTH and FLUX columns.

  • filters_root — root directory for filter response files (optional).

  • filters — defines the filters. Each entry requires:

    • magsys — magnitude system (ab, vega, bd17, or a custom standard name).

    • magzero — magnitude offset.

    • path — path to the filter response file.

    • lam_unit — wavelength unit in the response file. Default is Angstroms; set to nm or micron if needed.

Automatic filter dropping

The wavelength range covered by the model depends on which pre-trained model you use (see Built-in models). Filters are automatically dropped for individual SNe when they fall outside the rest-frame wavelength range at the source redshift. The cut-off wavelengths for each filter are defined as the wavelength where the filter response first drops below 1 per cent of its maximum value.