Cranlogs Dashboard

CRAN packages

I mentioned in an earlier post that I like to keep track of new and updated packages

On average, a new package is provided to CRAN every 5 hours and updates are even more numerous

CRAN provides an availablity table and there is also the informative CRANberries site plus twitter announcements but it is still hard to keep track of all the new developments

So I created a dashboard which is updated daily and provides a taster of the new. Because it has interactive inputs (e.g. selection of package name ) using shiny runtime, it takes somewhat longer to run than if there were htmlwidget inputs available

Here is an example of one of the outputs comparing downloads of a selection of interactive packages over time. It is a large file so is limited to the selected packages

library(feather)
library(plotly)
library(tidyverse)

# The choices can be amended in the actual app
choices <-
c("rbokeh", "ggvis", "plotly", "iplots", "rggobi", "googlevis")
# Pre prepared limited data of above choices
df <- read_feather("data/selectCranLogs.feather")


df %>%
filter(package %in% choices) %>%
plot_ly(x = ~ date,
y = ~ count) %>%
add_markers(color = ~ package) %>%
layout(
title = "Daily Downloads from CRAN",
xaxis = list(title = ""),
yaxis = list(title = "Count")
) %>%
config(displayModeBar = F, showLink = F)

Needs work!

However, it does show that the intensive work on the plotly package - covered in a book by the author of the package, Carson Sievert - has contributed to its popularity which spiked at the end of the period with a new release

Share Comments
comments powered by Disqus