Data Tip: Assignment with the ’T’ Pipe

dplyr

magrittr

Author

Tinashe M. Tapera

Published

January 17, 2019

So today, I found myself in an interesting situation wherein I was writing out an R object to disk, but needed to keep the file’s name and path in session so I could do something with it later. Typically, doing that would look like this:

# first, make some data, to write to disk
mydat <- data(iris) 
fname <- "iris.csv"
write.csv(x = mydat, fname)

# do some coding
# do some coding
# do some coding

# later on, reference the file using the fname variable
file.exists(fname)
file.remove(fname)

For some, this might be perfectly fine, but I found another interesting way of doing this using magrittr’s “Tee” pipe (%T>%):

shhh <- suppressPackageStartupMessages
shhh(library(dplyr, quietly = TRUE))
shhh(library(magrittr, quietly = TRUE))

fname2 <- "iris2.csv" %T>%
  write.csv(mydat, .)

file.exists(fname2)
file.remove(fname2)

Just an interesting way to assign an object and send the LHS of the pipe along the chain in one go.

---
title: 'Data Tip: Assignment with the ’T’ Pipe'
author: Tinashe M. Tapera
date: '2019-01-17'
slug: data-tip-assignment-with-the-t-pipe
categories:
  - dplyr
  - magrittr
tags:
  - dplyr
  - magrittr
# comments: no
showcomments: yes
showpagemeta: yes
execute: 
  eval: false
---

So today, I found myself in an interesting situation wherein I was writing out an R object to disk, but needed to keep the file's name and path in session so I could do something with it later. Typically, doing that would look like this:

```{r}
# first, make some data, to write to disk
mydat <- data(iris) 
fname <- "iris.csv"
write.csv(x = mydat, fname)

# do some coding
# do some coding
# do some coding

# later on, reference the file using the fname variable
file.exists(fname)
file.remove(fname)
```

For some, this might be perfectly fine, but I found another interesting way of doing this using `magrittr`'s "Tee" pipe (`%T>%`):

```{r, warning=FALSE}
shhh <- suppressPackageStartupMessages
shhh(library(dplyr, quietly = TRUE))
shhh(library(magrittr, quietly = TRUE))

fname2 <- "iris2.csv" %T>%
  write.csv(mydat, .)

file.exists(fname2)
file.remove(fname2)
```

Just an interesting way to assign an object and send the LHS of the pipe along the chain in one go.