Injection frames
================

Minke is able to create frame files which contain injections, which can then be used to do things like characterising parameter estimation algorithms.

It is possible to do this by writing a script.
However, Minke also provides a function to do this, and a commandline interface.
It is also possible to produce injection frames as a step in an asimov workflow.

Script
^^^^^^

::

  from minke.noise import AdvancedLIGO
  import astropy.units as u

  from minke.models.cbc import IMRPhenomPv2
  from minke.detector import AdvancedLIGOHanford

  noise = AdvancedLIGO()

  noise_ts = noise.time_series(duration=4, sample_rate=16384, epoch=998)

  model = IMRPhenomPv2()

  parameters = {"mass_ratio": 0.7, "total_mass": 100*u.solMass, "luminosity_distance": 100*u.megaparsec, "gpstime": 1000}

  data = model.time_domain(parameters, times=noise_ts.times)
  f = data['plus'].plot()
  f.savefig("waveform.png")

  detector = AdvancedLIGOHanford()

  projected = data.project(detector,
	       ra=1, dec=0.5,
	       iota=0.4,
	       phi_0=0,
	       psi=0
	       )
  injection = (noise_ts + projected)
  f = injection.plot()
  f.savefig("projected_injection.png")
  injection.write("test_injection.gwf", format="gwf")


.. image:: images/injections/projected_injection_script.png
   

The `make_injections` Functions
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

::

   from minke.injection import make_injection
   from minke.noise import AdvancedLIGO
   from minke.models.cbc import IMRPhenomPv2
   from minke.detector import AdvancedLIGOHanford

   import astropy.units as u


   detectors = {"AdvancedLIGOHanford": "AdvancedLIGO"}

   parameters = {"mass_ratio": 0.7,
		 "total_mass": 100*u.solMass,
		 "luminosity_distance": 100*u.megaparsec,
		 "ra": 1,
		 "dec": 0.5,
		 "iota": 0.4,
		 "phi_0": 0,
		 "psi": 0,
		 "gpstime": 1000}

   injection = make_injection(detectors=detectors, injection_parameters=parameters, duration=4, sample_rate=4096, epoch=998)['H1']

   f = injection.plot()
   f.savefig("injection_script.png")

.. image:: images/injections/projected_injection_function.png

We can also produce injections into zero noise:

::

   from minke.injection import make_injection_zero_noise
   from minke.noise import AdvancedLIGO
   from minke.models.cbc import IMRPhenomPv2
   from minke.detector import AdvancedLIGOHanford

   import astropy.units as u


   detectors = {"AdvancedLIGOHanford": "AdvancedLIGO"}

   parameters = {"mass_ratio": 0.7,
		 "total_mass": 100*u.solMass,
		 "luminosity_distance": 100*u.megaparsec,
		 "ra": 1,
		 "dec": 0.5,
		 "iota": 0.4,
		 "phi_0": 0,
		 "psi": 0,
		 "gpstime": 1000}

   injection = make_injection_zero_noise(detectors=detectors, injection_parameters=parameters, duration=4, sample_rate=16384, epoch=998)['H1']

   f = injection.plot()
   f.savefig("injection_function_zero.png")


.. image:: images/injections/projected_injection_zero.png   


	   
The Command-line interface
^^^^^^^^^^^^^^^^^^^^^^^^^^

To use the command line interface you'll need to create a YAML-formatted configuration file.

For example ::
  
   injection:
     duration: 4
     sample_rate: 4096
     epoch: 998
     channel: Injection
     parameters:
       luminosity_distance: 400
       m1: 35
       m2: 30
     waveform: IMRPhenomPv2
     interferometers:
       H1: AdvancedLIGOHanford
       L1: AdvancedLIGOLivingston
     psds:
       H1: AdvancedLIGO
       L1: AdvancedLIGO

Save this file as ``settings.yml``.
You can then create the injection files by running ``minke injection --settings settings.yml``       

As a step in an asimov workflow
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^