Metadata-Version: 2.2
Name: vaeesr
Version: 0.0.3
Summary: Using the latent space of a variational autoencoder to perform symbolic regression by sampling equations.
Author: Lisa Artmann
Author-email: lisaartmann01@gmail.com
Keywords: Symbolic Regression,Equation Discovery,Variational Autoencoder
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Science/Research
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Education
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: MacOS :: MacOS X
Classifier: Operating System :: Microsoft :: Windows
Classifier: License :: OSI Approved :: MIT License
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
License-File: LICENSE
Requires-Dist: equation-tree
Requires-Dist: numpy
Requires-Dist: tqdm
Requires-Dist: torch
Requires-Dist: pandas
Requires-Dist: sympy
Requires-Dist: scipy
Requires-Dist: scikit-learn
Requires-Dist: pyro-ppl
Dynamic: author
Dynamic: author-email
Dynamic: classifier
Dynamic: description
Dynamic: keywords
Dynamic: requires-dist
Dynamic: summary


Variational Autoencoder Embeddings Symbolic Regression (VAEESR)

This Python package performs Symbolic Regression by creating an Embedding where semantically similar equations are close to each other and it uses MCMC sampling to find the equation that is the closest to the observed data. Therefore, it uses three main functions: 

1. `create_dataset` which creates a customizable dataset that can be adjusted for the specific problem at hand. The main parameters are: The x_values for which the functions are evaluated, the range of constants, the maximum tree depth, the possible operators and functions, and the total number of equations in the dataset.

2. `create_autoencoder` which trains an autoencoder with the dataset. Some hyperparameters can be adjusted as well. 

3. `perform_MCMC` which performs the symbolic regression by sampling equations from the autoencoder embedding. 
