Spectre Normalization Tool (SNT)
git clone [email protected]:iastro-pt/Spectra-Normalization-Tool.git
cd Spectra-Normalization-Tool
pip install .
See the jupyter notebook in the docs/quickstart.ipynb folder
The algorithm allows some configuration/tuning of the initial parameters.
Either do:
from SNT.utils.SNT_configs import construct_SNT_configs
confs = construct_SNT_configs()
confs.print_table_of_descriptions()
Or check the full table that is provided here:
Name | description |
---|---|
remove_n_first | removes the first n points in the spectra (only if first points are not useful!) otherwise set to 0 |
radius_min | min alpha shape radius |
radius_max | max alpha shape radius (should be at least the size of the largest gap) |
max_vicinity | required number of values between adjacent maxima |
stretching | normalization parameter |
use_RIC | use RIC to avoid large 'dips' in the continuum (see documentation) |
interp | interpolation type |
use_denoise | for noisy spectra use the average of the flux value around the maximum |
usefilter | use savgol filter to smooth spectra |
nu | exponent of the computed penalty (see documentation) |
niter_peaks_remove | number of iterations to remove sharpest peaks before interpolation |
denoising_distance | number of points to calculate the average around a maximum if use_denoise is True, useful for noisy spectra |
Diogo Marques André M. Silva