README.md

In Thdegraaff/metareg: metareg a package for a meta analysis on a complete subset of regressions

metareg

Overview

metareg gives a complete subset of regression analyses by varying the number of independent variables. At the moment, there are only the following two functions (checked and worked) in this package. One to do the simulation of regressions and the other to analyse the results:

• result <- meta_ols(mpg ~., data = mtcars, k = 4)
• analyse_meta(result[[1]], result[[2]])

Installation

To install the package you need to have the devtools package installed in R. Then do:

library(devtools)
install_github("thdegraaff/metareg")

Once installation is completed, the package can be loaded.

library(metareg)

Note

Note that this package at the moment is very much work in progress and that the data it uses is the well-know mtcars dataset, which is built in in R. The data that will be used later is proprietary.

What is still to be incorporated (...at least)?

• Create a function that displays the distribution of (standardized) coefficients in a sort of funnel plot (or the forest plot of the meta package)
• Know how, but need to create the function
• Create a function that analyses the impact of the presence of control variables by using fixed effects on every variable
• Create a function that perform fixed effects estimation (with lfe) instead of OLS
• Update the vignette
• Not so important for now, but useful for documentation and first draft of the report

Getting help

If you need assistance using regrobustsim, you can get in touch by emailing [e-mail].

Thdegraaff/metareg documentation built on May 6, 2019, 8:02 p.m.