fastecon

fastecon is a high-performance R package for fast, scalable econometric analysis.
It is designed to simplify and accelerate common tasks in empirical research, including rolling window regressions, clustered standard errors, and table formatting for publication.

Features

• Fast rolling OLS with matrix caching
• Optional parallel computation
• Clean, minimal API
• Output formatted for LaTeX tables
• Modular design for integration into larger workflows

Installation

# Install from GitHub 
devtools::install_github("MengqiLi0000/fastecon")

Functions

1. rollreg_fast()

Purpose: Run fast rolling window OLS for time-series or panel data. Uses cached matrix updates.

Usage:

result <- rollreg_fast(data, y = "y", x = c("x1", "x2"), time = "date", window = 60, step = 1)

2. etable_latex()

Purpose: Create clean LaTeX regression tables from model output

Usage:

etable_latex(list(lm(y ~ x1, data = df), lm(y ~ x1 + x2, data = df)))

3. clustboot_fast()

Purpose: Run clustered bootstrap for linear regression. Supports one-way clustering and parallelism.

Usage:

clustboot_fast(data, formula = y ~ x1 + x2, cluster = "firm", B = 200, seed = 123)

4. feols_fast()

Purpose: Estimate fixed effects regression (1-way or 2-way) using within-transformation.

Usage:

feols_fast(data, y = "y", x = c("x1", "x2"), fe = c("firm", "year"))

5. forecast_fast()

Purpose: Make 1-step-ahead predictions using a model or coefficient vector.

Usage:

forecast_fast(model, newdata = df[1, ], xvars = c("x1", "x2"), ci = TRUE)