Module filters

source code

Functions relevant for photometric calibration

Table of contents:

1. Available response functions
2. Adding filters on the fly
   • Defining a new filter
   • Temporarily modifying an existing filter
3. Adding filters permanently

Section 1. Available response functions

Short list of available systems:

>>> responses = list_response()
>>> systems = [response.split('.')[0] for response in responses]
>>> set_responses = sorted(set([response.split('.')[0] for response in systems]))
>>> for i,resp in enumerate(set_responses):
...    print '%10s (%3d filters)'%(resp,systems.count(resp))
     2MASS (  3 filters)
    ACSHRC ( 17 filters)
    ACSSBC (  6 filters)
    ACSWFC ( 12 filters)
     AKARI ( 13 filters)
       ANS (  6 filters)
      APEX (  1 filters)
     ARGUE (  3 filters)
    BESSEL (  6 filters)
   BESSELL (  6 filters)
     COROT (  2 filters)
   COUSINS (  3 filters)
       DDO (  7 filters)
     DENIS (  3 filters)
     DIRBE ( 10 filters)
   EEV4280 (  1 filters)
     ESOIR ( 10 filters)
      GAIA (  4 filters)
     GALEX (  2 filters)
    GENEVA (  7 filters)
 HIPPARCOS (  1 filters)
     IPHAS (  3 filters)
      IRAC (  4 filters)
      IRAS (  4 filters)
    ISOCAM ( 21 filters)
   JOHNSON ( 25 filters)
    KEPLER ( 43 filters)
      KRON (  2 filters)
   LANDOLT (  6 filters)
      MIPS (  3 filters)
      MOST (  1 filters)
       MSX (  6 filters)
    NARROW (  1 filters)
    NICMOS (  6 filters)
      OAO2 ( 12 filters)
      PACS (  3 filters)
      SAAO ( 13 filters)
     SCUBA (  6 filters)
      SDSS ( 10 filters)
     SLOAN (  2 filters)
     SPIRE (  3 filters)
  STEBBINS (  6 filters)
   STISCCD (  2 filters)
   STISFUV (  4 filters)
   STISNUV (  7 filters)
 STROMGREN (  6 filters)
       TD1 (  4 filters)
     TYCHO (  2 filters)
    TYCHO2 (  2 filters)
  ULTRACAM (  5 filters)
    USNOB1 (  2 filters)
      UVEX (  5 filters)
   VILNIUS (  7 filters)
     VISIR ( 13 filters)
  WALRAVEN (  5 filters)
     WFCAM (  5 filters)
     WFPC2 ( 21 filters)
      WISE (  4 filters)
      WOOD ( 12 filters)

Plots of all passbands of all systems:

]include figure]]ivs_sed_filters_2MASS.png]

]include figure]]ivs_sed_filters_ACSHRC.png]

]include figure]]ivs_sed_filters_ACSSBC.png]

]include figure]]ivs_sed_filters_ACSWFC.png]

]include figure]]ivs_sed_filters_AKARI.png]

]include figure]]ivs_sed_filters_ANS.png]

]include figure]]ivs_sed_filters_APEX.png]

]include figure]]ivs_sed_filters_ARGUE.png]

]include figure]]ivs_sed_filters_BESSEL.png]

]include figure]]ivs_sed_filters_BESSELL.png]

]include figure]]ivs_sed_filters_COROT.png]

]include figure]]ivs_sed_filters_COUSINS.png]

]include figure]]ivs_sed_filters_DDO.png]

]include figure]]ivs_sed_filters_DENIS.png]

]include figure]]ivs_sed_filters_DIRBE.png]

]include figure]]ivs_sed_filters_ESOIR.png]

]include figure]]ivs_sed_filters_EEV4280.png]

]include figure]]ivs_sed_filters_GAIA.png]

]include figure]]ivs_sed_filters_GALEX.png]

]include figure]]ivs_sed_filters_GENEVA.png]

]include figure]]ivs_sed_filters_HIPPARCOS.png]

]include figure]]ivs_sed_filters_IPHAS.png]

]include figure]]ivs_sed_filters_IRAC.png]

]include figure]]ivs_sed_filters_IRAS.png]

]include figure]]ivs_sed_filters_ISOCAM.png]

]include figure]]ivs_sed_filters_JOHNSON.png]

]include figure]]ivs_sed_filters_KEPLER.png]

]include figure]]ivs_sed_filters_KRON.png]

]include figure]]ivs_sed_filters_LANDOLT.png]

]include figure]]ivs_sed_filters_MIPS.png]

]include figure]]ivs_sed_filters_MOST.png]

]include figure]]ivs_sed_filters_MSX.png]

]include figure]]ivs_sed_filters_NARROW.png]

]include figure]]ivs_sed_filters_NICMOS.png]

]include figure]]ivs_sed_filters_OAO2.png]

]include figure]]ivs_sed_filters_PACS.png]

]include figure]]ivs_sed_filters_SAAO.png]

]include figure]]ivs_sed_filters_SCUBA.png]

]include figure]]ivs_sed_filters_SDSS.png]

]include figure]]ivs_sed_filters_SLOAN.png]

]include figure]]ivs_sed_filters_SPIRE.png]

]include figure]]ivs_sed_filters_STEBBINS.png]

]include figure]]ivs_sed_filters_STISCCD.png]

]include figure]]ivs_sed_filters_STISFUV.png]

]include figure]]ivs_sed_filters_STISNUV.png]

]include figure]]ivs_sed_filters_STROMGREN.png]

]include figure]]ivs_sed_filters_TD1.png]

]include figure]]ivs_sed_filters_TYCHO.png]

]include figure]]ivs_sed_filters_TYCHO2.png]

]include figure]]ivs_sed_filters_ULTRACAM.png]

]include figure]]ivs_sed_filters_USNOB1.png]

]include figure]]ivs_sed_filters_UVEX.png]

]include figure]]ivs_sed_filters_VILNIUS.png]

]include figure]]ivs_sed_filters_VISIR.png]

]include figure]]ivs_sed_filters_WALRAVEN.png]

]include figure]]ivs_sed_filters_WFPC2.png]

]include figure]]ivs_sed_filters_WISE.png]

]include figure]]ivs_sed_filters_WOOD.png]

Section 2: Adding filters on the fly

Section 2.1: Defining a new filter

You can add custom filters on the fly using add_custom_filter. In this example we add a weird-looking filter and check the definition of Flambda and Fnu and its relation to the effective wavelength of a passband:

Prerequisites: some modules that come in handy:

>>> from ivs.sigproc import funclib
>>> from ivs.sed import model
>>> from ivs.units import conversions

First, we'll define a double peakd Gaussian profile on the wavelength grid of the WISE.W3 response curve:

>>> wave = get_response('WISE.W3')[0]
>>> trans = funclib.evaluate('gauss',wave,[1.5,76000.,10000.,0.])
>>> trans+= funclib.evaluate('gauss',wave,[1.0,160000.,25000.,0.])

This is what it looks like:

>>> p = pl.figure()
>>> p = pl.plot(wave/1e4,trans,'k-')
>>> p = pl.xlabel("Wavelength [micron]")
>>> p = pl.ylabel("Transmission [arbitrary units]")

]include figure]]ivs_sed_filters_weird_trans.png]

We can add this filter to the list of predefined filters in the following way (for the doctests to work, we have to do a little work around and call filters via that module, this is not needed in a normal workflow):

>>> model.filters.add_custom_filter(wave,trans,photband='LAMBDA.CCD',type='CCD')
>>> model.filters.add_custom_filter(wave,trans,photband='LAMBDA.BOL',type='BOL')

Note that we add the filter twice, once assuming that it is mounted on a bolometer, and once on a CCD device. We'll call the filter LAMBDA.CCD and LAMBDA.BOL. From now on, they are available within functions as get_info and get_response. For example, what is the effective (actually pivot) wavelength?

>>> effwave_ccd = model.filters.eff_wave('LAMBDA.CCD')
>>> effwave_bol = model.filters.eff_wave('LAMBDA.BOL')
>>> print(effwave_ccd,effwave_bol)
(119263.54911400242, 102544.27931275869)

Let's do some synthetic photometry now. Suppose we have a black body atmosphere:

>>> bb = model.blackbody(wave,5777.)

We now calculate the synthetic flux, assuming the CCD and BOL device. We compute the synthetic flux both in Flambda and Fnu:

>>> flam_ccd,flam_bol = model.synthetic_flux(wave,bb,['LAMBDA.CCD','LAMBDA.BOL'])
>>> fnu_ccd,fnu_bol = model.synthetic_flux(wave,bb,['LAMBDA.CCD','LAMBDA.BOL'],units=['FNU','FNU'])

You can see that the fluxes can be quite different when you assume photon or energy counting devices!

>>> flam_ccd,flam_bol
(897.68536911320564, 1495.248213834755)
>>> fnu_ccd,fnu_bol
(4.2591095543803019e-06, 5.2446332430111098e-06)

Can we now readily convert Flambda to Fnu with assuming the pivot wavelength?

>>> fnu_fromflam_ccd = conversions.convert('erg/s/cm2/AA','erg/s/cm2/Hz',flam_ccd,wave=(effwave_ccd,'A'))
>>> fnu_fromflam_bol = conversions.convert('erg/s/cm2/AA','erg/s/cm2/Hz',flam_bol,wave=(effwave_bol,'A'))

Which is equivalent with:

>>> fnu_fromflam_ccd = conversions.convert('erg/s/cm2/AA','erg/s/cm2/Hz',flam_ccd,photband='LAMBDA.CCD')
>>> fnu_fromflam_bol = conversions.convert('erg/s/cm2/AA','erg/s/cm2/Hz',flam_bol,photband='LAMBDA.BOL')

Apparently, with the definition of pivot wavelength, you can safely convert from Fnu to Flambda:

>>> print(fnu_ccd,fnu_fromflam_ccd)
(4.2591095543803019e-06, 4.259110447428463e-06)
>>> print(fnu_bol,fnu_fromflam_bol)
(5.2446332430111098e-06, 5.2446373530017525e-06)

The slight difference you see is numerical.

Section 2.2: Temporarily modifying an existing filter

Under usual conditions, you are prohibited from overwriting an existing predefined response curve. That is, if you try to add_custom_filter with a photband that already exists as a file, a ValueError will be raised (this is not the case for a custom defined filter, which you can overwrite without problems!). If, for testing purposes, you want to use another definition of a predefined response curve, you need to set force=True in add_custom_filter, and then call

>>> set_prefer_file(False)

To reset and use the original definitions again, do

>>> set_prefer_file(True)

Section 3.: Adding filters permanently

Add a new response curve file to the ivs/sed/filters directory. The file should contain two columns, the first column is the wavelength in angstrom, the second column is the transmission curve. The units of the later are not important.

Then, call update_info. The contents of zeropoints.dat will automatically be updated. Make sure to add any additional information on the new filters manually in that file (e.g. is t a CCD or bolometer, what are the zeropoint magnitudes etc).

>>> p = pl.figure()
>>> for band in ['J','H','KS']:
...    p = pl.plot(*get_response('2MASS.%s'%(band)))

Add a custom filter to the set of predefined filters.

Extra keywords are:
    'eff_wave', 'type',
    'vegamag', 'vegamag_lit',
    'ABmag', 'ABmag_lit',
    'STmag', 'STmag_lit',
    'Flam0', 'Flam0_units', 'Flam0_lit',
    'Fnu0', 'Fnu0_units', 'Fnu0_lit',
    'source'
    
default C{type} is 'CCD'.
default C{photband} is 'CUSTOM.FILTER'

@param wave: wavelength (angstrom)
@type wave: ndarray
@param response: response
@type response: ndarray
@param photband: photometric passband
@type photband: str ('SYSTEM.FILTER')

>>> bands, func = make_color('JOHNSON.B-V')
>>> print(bands)
('JOHNSON.B', 'JOHNSON.V')
>>> print(func(2,3.))
0.666666666667

>>> eff_wave('2MASS.J')
12393.093155655277