Construction of stellar templates¶

The construction of Stellar templates is handled through the StellarModel interface, which will create independent stellar templates for each sub-Instrument that we have available in the DataClass. The construction of the stellar templates follows the framework described in Silva et al, 2022, and is shematized in here:

In [1]:

from pathlib import Path

from ASTRA.data_objects import DataClassManager
from ASTRA.data_objects.DataClass import DataClass
from ASTRA.Instruments import ESPRESSO
from ASTRA.template_creation.StellarModel import StellarModel

storage_path = Path("../../tmp")
manager = DataClassManager()
manager.start()

d: DataClass = manager.DataClass(
    list(Path("/home/amiguel/spectra_collection/ESPRESSO/proxima").glob("*.fits")),
    storage_path=storage_path,
    instrument=ESPRESSO,
)

stell = StellarModel(
    storage_path,
    user_configs={},
)
stell.Generate_Model(
    dataClass=d,
    template_configs={},
    force_computation=True,
)

from pathlib import Path from ASTRA.data_objects import DataClassManager from ASTRA.data_objects.DataClass import DataClass from ASTRA.Instruments import ESPRESSO from ASTRA.template_creation.StellarModel import StellarModel storage_path = Path("../../tmp") manager = DataClassManager() manager.start() d: DataClass = manager.DataClass( list(Path("/home/amiguel/spectra_collection/ESPRESSO/proxima").glob("*.fits")), storage_path=storage_path, instrument=ESPRESSO, ) stell = StellarModel( storage_path, user_configs={}, ) stell.Generate_Model( dataClass=d, template_configs={}, force_computation=True, )

/home/amiguel/development/ASTRA/.venv/lib/python3.12/site-packages/numpy/lib/nanfunctions.py:1215: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)

Accessing the templates¶

After construction of the Stellar template, we can access them independently for each sub-Instrument:

In [2]:

stellar_template = stell.request_data("ESPRESSO19")

stellar_template = stell.request_data("ESPRESSO19")

This stellar_template object will have the same spectra-interfacing as the Frames:

In [12]:

_ = stellar_template.plot_spectra(order=50)

_ = stellar_template.plot_spectra(order=50)

---

Load stellar template without interfacing with DataClass¶

It is possible to directly load the stellar template without needing a valid DataClass object. However, doing this is not as direct as using a StellarModel, needing to manually load each sub-Instrument worth of data:

In [17]:

from ASTRA.template_creation.stellar_templates.sum_stellar import SumStellar

stell = StellarModel(
    storage_path,
    user_configs={},
)

stell.Generate_Model(
    dataClass=None,
    template_configs={
        "NUMBER_WORKERS": 1,
        "NORMALIZE_SPECTRA": True,
    },
    force_computation=False,
)

loaded_stellar_template = stell.request_data("ESPRESSO19")

_ = loaded_stellar_template.plot_spectra(order=50)

from ASTRA.template_creation.stellar_templates.sum_stellar import SumStellar stell = StellarModel( storage_path, user_configs={}, ) stell.Generate_Model( dataClass=None, template_configs={ "NUMBER_WORKERS": 1, "NORMALIZE_SPECTRA": True, }, force_computation=False, ) loaded_stellar_template = stell.request_data("ESPRESSO19") _ = loaded_stellar_template.plot_spectra(order=50)