The resulting model combines existent photometric, parallax, and chemical abundance of Lithium data sets of stars belonging to stellar open clusters to infer its age distribution through modern and robust artificial intelligence methods. A Neural Network is trained given a grid of pre-calculated BT-Settl models to interpolate the spectral energy distributions of stars, working as a black-box interpolator in the model. The Bayesian hierarchical model not only facilitates simultaneous inference of star-level parameters but also offers an elegant framework for effectively pooling open cluster information and propagating uncertainty. Markov Chain Monte Carlo techniques allow us to sample the posterior distribution using the Hamiltonian Monte Carlo algorithm.
We recommend using Anaconda (or Miniforge) to install Python on your local machine, which allows for packages to be installed using its conda utility.
Once you have installed one of the above, our software can be installed into a new conda environment as follows*:
# Create conda environment called "envname"
conda create -c conda-forge -n envname python=3.10
# Activate the new environment
conda activate envname
# Finally install biosc
pip install git+https://github.com/franciscopalmeromoya/biosc.gitAfter completing the installation process, you can confirm its success by running the following command:
python3 main.pyThis instructions have been tested on macOS and Ubuntu and have been confirmed to work smoothly. If you encounter any issues or have questions, please do not hesitate to contact us.
If you are developing, please consider install the package in "editable" mode with pip install -e. This allows you to make changes to the source code, and the changes will immediately affect the installed package without the need for reinstalling it.
One of the potential issues comes from the PyTensor installation. The library, required for PyMC, relies on low-level routines such as BLAS for performing common linear algebra operations. The low-level nature of such routines makes it inconvenient for a seamlessly installation in every device. If you find any problem related to pytensor, please consider manually install PyMC from its repository:
pip install git+https://github.com/pymc-devs/pymc.gitPlease, note that you can also change the requirements.txt file to install it as you did following the installation instructions.
To use the Bayesian model, follow these steps:
- Data Preprocessing: Use the
preprocessingmodule to prepare your data, including parallax, photometric, and lithium measurements.
from biosc.preprocessing import Preprocessing
# Example: Load and preprocess data
prep = Preprocessing('Pleiades_GDR3+2MASS+PanSTARRS1+EW_Li.csv', sortPho=False)
parallax_data = prep.get_parallax()
Li_data = prep.get_Li()
m_data = prep.get_magnitude(fillna='max')- Model Configuration and Compilation: Create an instance of the BayesianModel class and configure the model with priors and optional parameters.
from biosc.bhm import BayesianModel
# Example: Configure and compile the model
priors = {
'age': {'dist': 'uniform', 'lower': 0, 'upper': 200},
'distance': {'dist': 'normal', 'mu': 135, 'sigma': 20}
}
model = BayesianModel(parallax_data, m_data, Li_data)
model.compile(priors, POPho=False, POLi=True)- Sampling: Run the sampling process to obtain posterior distributions.
# Example: Sample from the posterior distribution
model.sample(chains=4)- Posterior Predictive Sampling (Optional): Generate samples from the posterior predictive distribution.
# Example: Sample from the posterior predictive distribution
model.sample_posterior_predictive()- Saving Inference Data: Save the inference data to a NetCDF file.
# Example: Save the inference data
model.save("output_filename.nc")- Visualization: Plot posterior distributions, trace plots, and other diagnostic plots.
# Example: Plot posterior distributions
model.plot_trace(var_names=['age', 'distance'])
# Example: Plot posterior predictive checks
fig, axs = plt.subplots(1, 3, figsize=(10, 5))
az.plot_ppc(model.idata, var_names = 'parallax', ax = axs[0])
az.plot_ppc(model.idata, var_names = 'flux', ax = axs[1])
az.plot_ppc(model.idata, var_names = 'Li', ax = axs[2])
plt.show()
# Example: Plot QQ for parallax
fig, ax = plt.subplots(figsize=(5, 5))
ax.set_ylabel('observed parallax [mas]')
ax.set_xlabel('parallax [mas]')
model.plot_QQ('parallax', fig, ax)
plt.show()This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.