Recent Posts

Also on R-Bloggers

More Posts

Today I was struggling with a relatively simple operation: unnest() from the tidyr package. What it’s supposed to do is pretty simple. When you have a data.frame where one or multiple columns are lists, you can unlist these columns while duplicating the information in other columns if the length of an element is larger than 1. library(tibble) df <- tibble( a = LETTERS[1:5], b = LETTERS[6:10], list_column = list(c(LETTERS[1:5]), "F", "G", "H", "I") ) df ## # A tibble: 5 x 3 ## a b list_column ## <chr> <chr> <list> ## 1 A F <chr [5]> ## 2 B G <chr [1]> ## 3 C H <chr [1]> ## 4 D I <chr [1]> ## 5 E J <chr [1]> library(tidyr) unnest(df, list_column) ## # A tibble: 9 x 3 ## a b list_column ## <chr> <chr> <chr> ## 1 A F A ## 2 A F B ## 3 A F C ## 4 A F D ## 5 A F E ## 6 B G F ## 7 C H G ## 8 D I H ## 9 E J I I came across this a lot while working on data from Twitter since individual tweets can contain multiple hashtags, mentions, URLs and so on, which is why they are stored in lists.


I’m happy to announce that rwhatsapp is now on CRAN. After being tested by users on GitHub for a year now, I decided it is time to make the package available to a wider audience. The goal of the package is to make working with ‘WhatsApp’ chat logs as easy as possible. ‘WhatsApp’ seems to become increasingly important not just as a messaging service but also as a social network—thanks to its group chat capabilities.


Today is Valentine’s Day. And since both I and my sweetheart are R enthusiasts, here is how to say “I love you” using a statistical programming language:


hrt_dat <- data.frame(t = seq(0, 2 * pi, by = 0.01)) %>%
  bind_rows(data.frame(t = rep(max(.$t), 300))) %>% 
  mutate(xhrt = 16 * sin(t) ^ 3,
         yhrt = 13 * cos(t) - 5 * cos(2 * t) - 2 * cos(3 * t) - cos(4 * t),
         frame = seq_along(t)) %>% 
  mutate(text = ifelse(frame > 300, "            J", "")) %>%
  mutate(text = ifelse(frame > 500, "A           J", text)) %>%
  mutate(text = ifelse(frame > 628, "A     +     J", text)) %>% 
  mutate(texty = 0, textx = 0)

ggplot(hrt_dat, aes(x = xhrt, y = yhrt)) +
  geom_line(colour = "#C8152B") +
  geom_polygon(fill = "#C8152B") +
  geom_text(aes(x = textx, y = texty, label = text), 
            size = 18, 
            colour = "white",
            vjust = "center") +
  theme_void() +


Some time ago, I saw a presentation by Wouter van Atteveldt who showed that wordclouds aren’t necessarily stupid. I was amazed since wordclouds were one of the first things I ever did in R and they are still often shown in introductions to text analysis. But the way they are mostly done is, in fact, not very informative. Because the position of the individual words in the cloud do not mean anything, the only information communicated is through the font size and sometimes font colour of the words.


For my PhD project, I want to use Supervised Machine Learning (SML) to replicate my manual coding efforts onto a larger data set. That means, however, that I need to put in some manual coding effort before the SML algorithms can do their magic! I used a number of programs already to analyse texts by hand, and they all come with their up- and downsides. A while ago I already coded articles in order to train an SML algorithm and did so having a PDF with the text opened on the left side of my screen and an Excel file with my category system on the right side.




An R package for working with WhatsApp data


LexisNexisTools. An R Package for Working with Files from ‘LexisNexis’


Work on rDNA with Philip Leifeld


Download my academic CV (last update: 09 September 2019).