import numpy as np
import ForMoSA.utils.spec as us
from ForMoSA.core.errors import ForMoSAError
from ForMoSA.core.enums import VsiniFunction
from ForMoSA.transform.observed import ObservedModel
from ForMoSA.observation.observation_base import Observation
[docs]
class ApplyPhysicsEffects:
'''
Apply physics effects to a model given parameters.
Notes
-----
Authors: Allan Denis
'''
@staticmethod
def _apply_rv(observed_model: ObservedModel, rv_value: float) -> ObservedModel:
'''
Apply radial velocity transformation.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel
rv_value : float
RV value
Returns
-------
observed_model : ObservedModel
Instance of class ObservedModel Doppler shifted
Notes
-----
Authors: Allan Denis
'''
observed_model.wave, observed_model.flux, observed_model.res = us.doppler_fct(observed_model.wave, observed_model.flux, observed_model.res, rv_value)
return observed_model
@staticmethod
def _apply_vsini(observed_model: ObservedModel, vsini_value: float, ld_value: float, vsini_function: VsiniFunction) -> ObservedModel:
'''
Apply v.sini transformation.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel
vsini_value : float
v.sini value
ld_value : float
limbd darkening value
vsini_function : VsiniFunction
Function used for the v.sini
Returns
-------
observed_model : ObservedModel
Instance of class ObservedModel rotationally broadened
Notes
-----
Authors: Allan Denis
'''
observed_model.flux, observed_model.res = us.vsini_fct(observed_model.wave, observed_model.flux, observed_model.res, ld_value, vsini_value, vsini_function.value)
return observed_model
@staticmethod
def _apply_reddening(observed_model: ObservedModel, av_value: float) -> np.ndarray:
'''
Apply redening transformation.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel
av_value : float
Extinction value
Returns
-------
observed_model : ObservedModel
Instance of class ObservedModel redenned
Notes
-----
Authors: Allan Denis
'''
observed_model.flux = us.reddening_fct(observed_model.wave, observed_model.flux, av_value)
return observed_model
@staticmethod
def _apply_cpd(observed_model: ObservedModel, bb_T_value: float, bb_R_value: float, d_value: float) -> np.ndarray:
'''
Apply CPD effect.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel
bb_T_value : float
Black body temperature value
bb_R_value : float
Black body radius value
d_value : float
Distance value
Returns
-------
(ObservedModel): Instance of class ObservedModel with a CPD effect
Notes
-----
Authors: Allan Denis
'''
observed_model.flux = us.bb_cpd_fct(observed_model.wave, observed_model.flux, d_value, bb_T_value, bb_R_value)
return observed_model
@staticmethod
def _apply_scaling(observed_model: ObservedModel, r_value: float, d_value: float) -> np.ndarray:
'''
Apply scaling.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel
r_value : float
Radius value
d_value : float
Distance value
Returns
-------
flux_transformed : np.ndarray
Scaled model
Notes
-----
Authors: Allan Denis
'''
observed_model.flux, ck = us.calc_ck(observed_model.flux, [], [], r_value, d_value)
observed_model.scaling = 'physical'
return observed_model
[docs]
class ApplyObservationEffects:
@staticmethod
def _apply_resolution_decreasing(observed_model: ObservedModel, obs: Observation) -> ObservedModel:
'''
Apply resolution decreasing.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel to transform
obs : Observation
Observation
Returns
-------
(ObservedModel): Instance of class ObservedModel degraded in resolution
Notes
-----
Authors: Allan Denis
'''
if obs.res.all() == 0:
return observed_model
if len(observed_model.wave) != len(obs.wave):
flux = us.resolution_decreasing(observed_model.wave, observed_model.flux, observed_model.res, obs.wave, obs.res)
return ObservedModel(obs.wave, flux, obs.res)
return observed_model
@staticmethod
def _apply_scaling(observed_model: ObservedModel, obs: Observation, bounds: tuple[float, float] = (-float('inf'), float('inf'))) -> ObservedModel:
'''
Apply analytical scaling between a model and observations.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel to transform
obs : Observation
Observation
bounds : tuple[float, float]
Bounds for the Least Squares
Returns
-------
(ObservedModel): Instance of class ObservedModel scaled to the data
Notes
-----
Authors: Allan Denis
'''
if observed_model.scaling == 'analytic':
observed_model.flux, ck = us.calc_ck(observed_model.flux, obs.flux, obs.err, 0, 0, analytic='yes', bounds = bounds)
return observed_model
@staticmethod
def _hc_modeling(observed_model: ObservedModel, obs: Observation, bounds: tuple[float, float] = (-float('inf'), float('inf'))) -> ObservedModel:
'''
High-contrast modeling: solve flux_obs = sum_i coeffs_i * components_i
where components_0 is model and components_i (i >= 1) is stellar speckles or systematics.
Parameters
----------
observed_model : ObservedModel
Instance of class ObservedModel to transform
obs : Observation
Observation
bounds : tuple[float, float]
Bounds for the least squares
Returns
-------
(ObservedModel): Instance of class ObservedModel high-contrast modelled
Notes
-----
Authors: Allan Denis
'''
# =================================
# Initial checks
# =================================
if not obs.hc_mode:
raise ForMoSAError('Your observation needs to be in high-contrast mode in order to be able to use high-contrast modeling' )
# =================================
# Consistency between model and obs
# =================================
wave_cont, res_cont = obs.wave_cont, obs.res_cont
wave, res = obs.wave, obs.res
if len(wave) != len(observed_model.wave):
raise ForMoSAError(f'Wrong length for model wavelength: {len(observed_model.wave)}. Expected the same length as the observations: {len(wave)}' )
# High-contrast modeling requires wave_cont and res_cont for continuum estimation.
# These must be set via the config (wav_cont / res_cont in [config_adapt]).
if wave_cont is None or wave_cont == 'NA':
raise ForMoSAError(
'wave_cont is not set on the observation. '
'High-contrast modeling requires wav_cont to be defined in the config file (e.g. wav_cont = 1.4, 1.8).'
)
if res_cont is None or res_cont == 'NA':
raise ForMoSAError(
'res_cont is not set on the observation. '
'High-contrast modeling requires res_cont to be defined in the config file (e.g. res_cont = 100).'
)
observed_model.flux_cont = us.continuum_estimate(wave, observed_model.flux * obs.transm, res, wave_cont, res_cont)
# =================================
# Build components for linear fit
# =================================
components, fixed_coeffs, bounds, labels = us.build_linear_components(
observed_model.flux,
transm=obs.transm,
flx_cont_obs=obs.flux_cont,
flx_cont_mod=observed_model.flux_cont,
star_flx_obs=obs.star_flux,
star_flx_cont_obs=obs.star_flux_cont,
system_obs=obs.system,
analytic='yes',
bounds=bounds
)
# =================================
# Solve linear model
# =================================
result = us.fit_linear_model(
components=components,
flx_obs=obs.flux,
err_obs=obs.err,
bounds=bounds,
fixed_coeffs=fixed_coeffs
)
observed_model.flux, observed_model.component = result['reconstructed'][0], np.sum(result['reconstructed'][1: len(components)], axis= 0)
return observed_model
