README.md

In ALanguillaume/tagnfuse: Implement tag and fuse workflow

tagnfuse

An R package to play with lists of data.frames.

Overview

The purpose of this R package is to transform the names of a list of data.frames into a column of each data.frame. That comes in handy for me in two contexts:

• When your data is spread among several files and the names of the files correspond to the levels of a variable of interest. For example, someone sends you the same dataset for different locations…
• When you want to compare different statistical models. Let’s say you want to plot different models AIC using ggplot. You simply need to put your models into a list with adequate names, then reshape the model objects using the broom package, tagnfuse the list and you are good to go.

Installation

install.packages("devtools")
devtools::install_github("ALanguillaume/tagnfuse", dependencies = TRUE)

Usage

library(tagnfuse)

Example using dummy data.

foo <- list(bonnie = data.frame(dollars = c(100, 200),
                                bullets = c(3, 4)),
            clyde = data.frame(dollars = c(150, 75),
                               bullets = c(7, 8)))
foo
#> $bonnie
#>   dollars bullets
#> 1     100       3
#> 2     200       4
#> 
#> $clyde
#>   dollars bullets
#> 1     150       7
#> 2      75       8

Tag…

foo %>% tag(label = "gangster")
#> $bonnie
#>   dollars bullets gangster
#> 1     100       3   bonnie
#> 2     200       4   bonnie
#> 
#> $clyde
#>   dollars bullets gangster
#> 1     150       7    clyde
#> 2      75       8    clyde

…and then fuse !

foo %>% tag(label = "gangster") %>% fuse()
#>   dollars bullets gangster
#> 1     100       3   bonnie
#> 2     200       4   bonnie
#> 3     150       7    clyde
#> 4      75       8    clyde

Or directly

foo %>% tagnfuse(label = "gangster")
#>   dollars bullets gangster
#> 1     100       3   bonnie
#> 2     200       4   bonnie
#> 3     150       7    clyde
#> 4      75       8    clyde

Statement of purpose

This was an exercice for me to learn R package development. One should not use this package in production code. dplyr::bind_rows() or data.table::rbindlist() are the state of the art tools for such operations.

foo %>% dplyr::bind_rows(.id = "gangster")
#>   gangster dollars bullets
#> 1   bonnie     100       3
#> 2   bonnie     200       4
#> 3    clyde     150       7
#> 4    clyde      75       8

foo %>% data.table::rbindlist(idcol = "gangster")
#>    gangster dollars bullets
#> 1:   bonnie     100       3
#> 2:   bonnie     200       4
#> 3:    clyde     150       7
#> 4:    clyde      75       8

ALanguillaume/tagnfuse documentation built on Feb. 2, 2020, 5:43 p.m.