9 Tidyverse
Tidyverse的昨天、今天、和明天
Hadley Wickham在2025-10-09写了一篇文章,题目是“A personal history of the tidyverse”(1)。

我们认真的阅读,并整理出一个读书笔记(全是引用),或理解为知识点。每一条都可能是“干货”。😎
tidyverse was named in 2016.
I was very lucky to have access to computers from a very early age (around 10), thanks to my dad, and this led to a general interest in computers and programming.

…my undergraduate statistics degree at the University of Auckland (New Zealand), the birthplace of R.
I started using R (v1.6.2) in 2003.
… my dad had done his PhD at Cornell University (the USA) (with only two years).
… to do my PhD (statistics, 2004 - 2008) … Iowa State University (ISU) (the USA)…
At the same time (PhD) I was reading The Grammar of Graphics (by Leland Wilkinson) …only implementation available at the time was very expensive, so I decided I’d have to go at creating my own in R. That led to ggplot and later ggplot2.

After graduating from ISU, I got a job at Rice University (2008 - 2012) … teaching “Introduction to Data Analysis” … year-to-year … led to the creation of the stringr (2009) and lubridate (2010) packages.
… the popularity of ggplot2 continued to rise … carve out time to work on it … not valued by my department.
During my time at Rice I had very little success with grants … I was fortunate to get a couple of small grants from BD (a medical technolgy company) and Google to continue my mork on plyr, reshape, and ggplot2.
I was developing quite a few packages … invest in tooling … led to the creation of the testthat (2009) and devtools (2011) packages … and maintenance of the roxygen2 package (2011).
In 2012, I left Rice (University) for RStudio (now Posit) … where the practice of software engineering was valued and I no longer needed to produce papers or find grant money.

my first few years at RStudio led to an explosion of new packages … The most important new package was dplyr.
I took over maintenance of the DBI and RSQLite packages …, created bigrquery, and forked RPostgres … worked on a range of other data sources including web scraping (rvest), Excel (readxl), ractangular text files (readr), SPSS/SAS/Stata (haven), and XML (xml2) … relied on tight integration with existing C libraries.
Around this time I started to become particularly well known in the R community.
… I don’t like: it (the name tidyverse) implies that everything outside the tidyverse is the messyverse.
… all the books I write have a dual production model: a free HTML version that I produce and a paid version produced by a commercial publisher.
R for Data Science proved to be extremely popular … has been translated into including Russia, Polish, Japanese, Chinese (traditional) and Chinese (simplified).
I created the tibble, an extention of the data frame …
I first implemented the pipe in dplyr in Oct 2013 and called it %.% … use magrittr (%>%) … with the maturity of the base pipe, the tidyverse is gradually moving away from %>% towards |>.
… theory is really important but experience taught us that few R users wanted to learn it.
I’m pretty sure the first hex logo was magrittr’s, designed by Stefan in December 2014.
This (tidyverse) wouldn’t have been possible without Posit (former RStudio), which has funded this work as part of its public benefit corporation mission. I feel tremendously lucky to work at an organization whose mission so closely aligns with my own.
The mission of the (tidyverse) team is broad and extends well beyond the tidyverse: we want to make R the best environment for doing data science … we love R, believe in it, and believe that you have to stay focused if you want to have an impact.
… testthat for unit testing … pkgdown to make package websites … roxygen2 to generate documentation.
Within the tidyverse team, the tidymodels team has a narrower mission: improving the tools data scientists need to do statistical modelling and machine learning.
Positron is a new integrated development environment (IDE) for data science, produced by the same team that created RStudio.
Looking ahead, I’m more excited than ever about the future of R and the the tidyverse.
I still love R, love programming, and most importantly, love working with this incredible community.