a software for Bayesian time-series econometrics applications
Project description
Alexandria
Alexandria is a Python package for Bayesian time-series econometrics applications. This is the first official release of the software. For its first release, Alexandria includes only the most basic model: the linear regression. However, it proposes a wide range of Bayesian linear regressions:
- maximum likelihood / OLS regression (non-Bayesian)
- simple Bayesian regression
- hierarchical (natural conjugate) Bayesian regression
- independent Bayesian regression with Gibbs sampling
- heteroscedastic Bayesian regression
- autocorrelated Bayesian regression
Alexandria is user-friendly and can be used from a simple Graphical User Inteface (GUI). More experienced users can also run the models directly from the Python console by using the model classes and methods.
===============================
Installing Alexandria
Alexandria can be installed from pip: pip install alexandria-python A local installation can also obtain by copy-pasting the folder containing the toolbox programmes. The folder can be downloaded from the project website or Github repo: https://alexandria-toolbox.github.io https://github.com/alexandria-toolbox
===============================
Getting started
Simple Python example:
imports
from alexandria.linear_regression import IndependentBayesianRegression from alexandria.datasets import data_sets as ds import numpy as np
load Taylor dataset, split as train/test
taylor_data = ds.load_taylor() y_train, X_train = taylor_data[:198,0], taylor_data[:198,1:] y_test, X_test = taylor_data[198:,0], taylor_data[198:,1:]
set prior mean and prior variance for the model
b = np.array([1.5, 0.5]) b_const = 1 V = np.array([0.01, 0.0025]) V_const = 0.01
create and train regression
br = IndependentBayesianRegression(endogenous=y_train, exogenous=X_train, constant=True, b_exogenous=b, V_exogenous=V, b_constant=b_const, V_constant=V_const) br.estimate()
get predictions on test sample, run forecast evaluation, display log score
estimates_forecasts = br.forecast(X_test, 0.95) br.forecast_evaluation(y_test) print('log score on test sample : ' + str(round(br.forecast_evaluation_criteria['log_score'], 2)))
===============================
Documentation
Complete manuals and user guides can be found on the project website and Github repo: https://alexandria-toolbox.github.io/ https://github.com/alexandria-toolbox
===============================
Contact
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for alexandria_python-0.0.2-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 2bb368e515bf6ea3e679b12f2f1edcc272b9c43c44ac0081fbeed216ea5a7ef5 |
|
MD5 | 0d874ad265f6d67adcfecec93110a423 |
|
BLAKE2b-256 | 1f562baf5f4647ca8097f7a2bc64db62b1c31f272e84877d5d64f53edac596aa |