Table of contents
Notebooks on Time Series (subscribe via RSS)
The pages below document certain specific topics usually related to time series analysis in Python. Usually they are Jupyter Notebooks, rendered into HTML for this website.
Note: all notebooks are kept for historical context, and so as time has passed some have been superseded by others. Notes have been added to indicate where this has happened.
2020
- Large dynamic factor models, forecasting, and nowcasting
- TVP-VAR, MCMC, and sparse simulation smoothing
- Chandrasekhar recursions in Statsmodels
2018
2017
- Implementing and estimating a simple Real Business Cycle (RBC) model
- Implementing and estimating a local level state space model
- Implementing and estimating an ARMA(1, 1) state space model
2016
2015
- Dynamic factors and coincident indices
- Bayesian state space estimation in Python via Metropolis-Hastings
- Estimating a Real Business Cycle DSGE Model by Maximum Likelihood in Python
- State space diagnostics
- Unobserved components
2014
- State space modeling in Python
- Implementing state space models for Statsmodels
- Kalman Filter Initialization - The Stationary Case
2013
- Bernoulli Trials in Python: Bayesian Estimation
- Bernoulli Trials in Python: Classical Estimation
-
Markov-switching - Filardo (1994) Time-Varying Transition Probabilities
(superseded by "Markov switching autoregression models")
-
Markov-Switching - Kim, Nelson, and Startz (1998) Three-state Variance Switching
(superseded by "Markov switching autoregression models")
-
Markov-switching - Hamilton (1989) Markov Switching Model of GNP
(superseded by "Markov switching autoregression models")
- SETAR Model Functionality
- Developing with Python
Example notebooks for Statsmodels in 2021
Statsmodels is a Python library that contains many poweful statistical tools, but there aren't always enough examples of how and when to use it, and, worse, older examples can become outdated. This set of example notebooks is intended to showcase some of the ways that Statsmodels can be used for data science and econometrics.