A library for Bayesian econometrics

Image 01

Welcome to the official Alexandria website. Alexandria is a toolbox dedicated to Bayesian time-series econometrics models, developed jointly in Python and Matlab.



On this website you will find an overview of the toolbox, along with links to download the software and its manuals for free.

Latest news: Alexandria V 0.1 is now released! It can be downloaded along with its manuals in the download sections of the website.

Technologies

Romain Legrand

Image 02

Skilled economist, econometrician, quantitative researcher and data scientist with a passion for modelling and quantitative methods. I am the creator, developer and owner of Alexandria.



I hold A PhD in economics from Cergy Paris University (2014) and a Master's degree in Big Data from Telecom Paris (2020). I have worked as an economist and econometrician for the European Central Bank during four years (2014-2018). There I developed the BEAR toolbox, which is the Matlab ancestor of Alexandria. I now work as a quantitative researchr for Qube Research and Technologies, one of the most succesful hedge funds in France.

More about me

For more information, you can download my resume or visit my Linkedin page.

Getting started

Python install: pip install alexandria-python

Then (simple Bayesian regression model):

from alexandria.linear_regression import SimpleBayesianRegression
from alexandria.datasets import data_sets as ds

# load Taylor rule toy dataset (federal funds rate, inflation, output gap)
taylor_data = ds.load_taylor()
y, X = taylor_data[:,0], taylor_data[:,1:]

# train the model with default settings
sbr = SimpleBayesianRegression(endogenous=y, exogenous=X)
sbr.estimate()

# get predictions, run forecast evaluation, display log score
estimates_forecasts = sbr.forecast(X, 0.95)
sbr.forecast_evaluation(y)
print(sbr.forecast_evaluation_criteria['log_score'])

Console output: 2.2505

Contact

You can contact me with the following form, or directly by sending an email to alexandria.toolbox@gmail.com.