fixedCV

Welcome! fixedCV is an R package that provides robust statistical inference for time series and other dependent data.

What does it do?

When working with correlated data (like time series), standard statistical methods can give misleading results. This package provides tools for:

• Robust hypothesis testing that accounts for unknown correlation structures
• Reliable confidence intervals for regression coefficients in dependent data
• Long-run variance estimation using state-of-the-art kernel methods

The package implements fixed-b critical values and multiple kernel-based estimators to ensure your statistical inferences are valid even when observations are correlated.

Installation

You can install fixedCV directly from GitHub using the devtools package:

# Install devtools if you haven't already
install.packages("devtools")

# Install fixedCV from GitHub
devtools::install_github("rpkgarcia/fixedCV")

Alternatively, you can use the remotes package:

# Install remotes if you haven't already
install.packages("remotes")

# Install fixedCV from GitHub
remotes::install_github("rpkgarcia/fixedCV")

Quick Start

Once installed, load the package and you're ready to go:

library(fixedCV)

# Fit a linear model
model <- lm(y ~ x, data = your_data)

# Get robust inference that accounts for autocorrelation
robust_results <- robust_lm(model)

The robust_lm() function automatically selects appropriate bandwidth parameters and provides robust standard errors, t-statistics, and p-values that are valid under general dependence structures.

Getting Help

For detailed documentation and examples:

# View package documentation
help(package = "fixedCV")

# See examples for the main function
?robust_lm

# Access the package vignette
vignette("fixedCV-vignette")

Questions or Issues?

If you encounter any problems or have questions, please open an issue on the GitHub repository.