def test_resolutions2ints_lists(resolutions, results):
"""Test transformation of resolutions to integer resolutions."""
assert results == utils.resolutions2ints(resolutions)
def test_resolutions2ints_fails_on_single(resolutions):
with pytest.raises(TypeError):
utils.resolutions2ints(resolutions)
def main(
startype: Sequence[str],
vsini: Sequence[float],
resolution: Sequence[str],
band: Sequence[str],
sample_rate: Optional[List[Union[int, float]]] = None,
noresample: bool = False,
unnormalized: bool = False,
org: bool = False,
replace: bool = False,
):
"""Run convolutions of NIR spectra for the range of given parameters.
Multiple values of startype, vsini, resolution, band, and sample_rate can
be provided.
Read files from eniric.paths["phoenix_dat"]"
Parameters
----------
startype: list of strings
vsini: list of floats
resolution: list of floats
band: list of strings
sample_rate: list of floats default=[3.0]
noresample: bool default=False
unnormalized: bool default=False
"""
normalize = not unnormalized
if sample_rate is None:
# Default sample rate
sample_rate = [3.0]
# Check the inputs are correct format. (lists)
for f_input, f_name in zip(
[startype, band, vsini, resolution, sample_rate],
["startype", "band", "vsini", "resolution", "sample_rate"],
):
if not isinstance(f_input, list):
print(f_name, type(f_input), type(f_name))
raise TypeError("Input {0} is not list".format(f_name))
# vsini, resolution, band and sample_rate can all be a series of values
vsini = [float(v) for v in vsini] # turn to floats
# Handle K in Resolution
res = resolutions2ints(resolution)
start_time = dt.now()
phoenix_path = eniric.paths["phoenix_dat"]
results_dir = eniric.paths["results"]
os.makedirs(results_dir, exist_ok=True)
resampled_dir = eniric.paths["resampled"]
os.makedirs(resampled_dir, exist_ok=True)
if not normalize:
norm_ = "_unnormalized"
else:
norm_ = ""
counter = 0
for star in startype:
spectrum_name = os.path.join(phoenix_path,
get_spectrum_name(star, org=org))
for b in band:
for vel in vsini:
for R in res:
for sample in sample_rate:
r = int(float(R) / 1000)
result_name = "Spectrum_{0}-PHOENIX-ACES_{1}band_vsini{2:.01f}_R{3}k{4}.dat".format(
star, b, vel, r, norm_)
if (os.path.exists(
os.path.join(results_dir, result_name))
and not replace):
print("Skipping convolution as {} already exists".
format(result_name))
continue
print("Name to be the result file", result_name)
convolve_spectra(
spectrum_name,
b,
vel,
R,
epsilon=0.6,
fwhm_lim=5.0,
num_procs=None,
results_dir=results_dir,
normalize=normalize,
output_name=result_name,
)
# Resample only the file just made
if not noresample:
resampler(
result_name,
results_dir=results_dir,
resampled_dir=resampled_dir,
sampling=sample,
)
counter += 1
print("Time to convolve {0:d} combinations = {1}".format(
counter,
dt.now() - start_time))
return 0