# Software and Code

## bootUR: R Package for Bootstrap Unit Root Tests

The R package bootUR, co-authored with Ines Wilms, is available on CRAN and GitHub. In this package we implement several bootstrap unit root, both for univariate and multivariate time series, including (but not limited to) all the unit root tests proposed in my papers. For details see the website or the accompanying JSS paper.

The package encompasses all code available below that relates to unit root testing in a coherent and user-friendly way. As a consequence, all bootstrap unit root testing codes below are not updated anymore, as I will only keep the package up to date.

Questions and comments on the bootUR package are welcome via my GitHub page (issues tab). To cite the package in your research, use the command citation("bootUR") in R.

## desla: R Package for Desparsified Lasso Inference for Time Series

The R package desla, developed by Robert Adamek, is available on CRAN and GitHub. This package allows for inference in high-dimensional time series models via the desparsified lasso as proposed in the papers "Lasso inference for high-dimensional time series" and "Local projection inference in high dimensions".

## specs: R Package for the Single-Equation Penalized Error-Correction Selector

The R package specs, developed by Etienne Wijler, is available on CRAN and GitHub. In this package the Single-Equation Penalized Error-Correction Selector (SPECS) from the paper "An automated approach towards sparse single-equation cointegration modelling" is implemented. The data from the paper are also included.

## HDGCvar: R Package for Granger Causality Testing in High-Dimensional Vector Autoregressive Models

The R package HDGCvar, developed by Luca Margaritella, is available on GitHub. This package allows for testing Granger causality in high-dimensional vector autoregressive models (VARs) as proposed in the paper "Granger causality testing in high-dimensional VARs: a Post-Double-Selection procedure" and "Inference in non-stationary high-dimensional VARs".

## TransmissionChannelAnalysis.jl: Julia Package for Estimating Dynamic Transmission Channels

The Julia package TransmissionChannelAnalysis.jl, developed by Enrico Wegner, is available on GitHub. This package allows for estimating transmission channels as proposed in the paper "Transmission Channel Analysis in Dynamic Models".

## Individual Codes (not in a package)

Code and data for replication of the results in the paper "Min(d)ing the President: A text analytic approach to measuring tax news".

R code and Python code for the trend analysis methods proposed in the paper "A statistical analysis of time trends in atmospheric ethane".

R code and datasets for the estimation of the high-dimensional state space model proposed in the paper "A dynamic factor model approach to incorporate Big Data in state space models for official statistics".

MATLAB code for residual bootstrap methods for VaR developed in the paper "A residual bootstrap for conditional Value-at-Risk".

R code for the AWB trend inference method developed in the paper "Autoregressive wild bootstrap inference for nonparametric trends".

MATLAB toolbox for the WIMP method developed in the paper "Inference for impulse responses under model uncertainty". The code and data in the toolbox allow for replication of the WIMP confidence bands for all the identification schemes used in the empirical study in the paper.

GAUSS code and R code for the bootstrap panel predictability tests developed in the paper "Robust block bootstrap panel predictability tests".

GAUSS code for the bootstrap ECM cointegration tests developed in the paper "A sieve bootstrap test for cointegration in a conditional error correction model".

### Legacy Codes (superseded by bootUR)

GAUSS code and R code for the modified wild bootstrap tests developed in the paper "A multivariate invariance principle for modified wild bootstrap methods with an application to unit root testing". The code includes the lag length selection methods developed in the paper "Lag length selection for unit root tests in the presence of nonstationary volatility".

GAUSS code and R code for the bootstrap panel predictability tests developed in the paper "Robust block bootstrap panel predictability tests".

GAUSS code and R code for the bootstrap sequential quantile tests developed in the paper "Bootstrap sequential tests to determine the order of integration of individual units in a time series panel".

GAUSS code for the bootstrap union tests developed in the paper "Bootstrap union tests for unit roots in the presence of nonstationary volatility".

GAUSS code for the bootstrap panel unit root tests developed in the paper "Cross-sectional dependence robust block bootstrap panel unit root tests".

### Notes

All codes are provided under a GPL 2 or later license.

The individual codes are fully functional but provided without any warranty whatsoever. I will also not update them except to fix bugs. The R code has been tested to give (almost) identical output to the Gauss output using a limited number of datasets. Questions on how to use those codes, or comments and suggestions on typos, improvements, etc. are welcome by e-mail. If you use the code in your research, an acknowledgement in the form of a reference to this website and a citation of the relevant papers is appreciated.

The zip-files with Gauss codes contain two files: a ‘.gss’ file with the original Gauss code, and a ‘.src’ file with modifications of that code such that it can be run in OxGauss. Typically the modifications, such as changing the random number generator, do not affect the functionality of the code; the exception is the MWB.src file which has a minor loss of functionality (see the file itself for more information).

R is available for free download here.

OxGauss is part of Ox; the Ox Console version is freely available for academic use and can be downloaded here.