wikipedia/example_analysis/pageview_example.R

   1 ### COVID-19 Digital Observatory
   2 ### 2020-03-28
   3 ###
   4 ### Minimal example analysis file using pageview data
   5
   6 library(tidyverse)
   7 library(scales)
   8
   9 ### Import and cleanup one datafile from the observatory
  10
  11 DataURL <-
  12     url("https://covid19.communitydata.science/datasets/wikipedia/digobs_covid19-wikipedia-enwiki_dailyviews-20200101.tsv")
  13
  14 views <-
  15     read.table(DataURL, sep="\t", header=TRUE, stringsAsFactors=FALSE)
  16
  17 ### Cleanup and do the grouping with functions from the Tidyverse
  18 ### (see https://www.tidyverse.org for more info)
  19
  20 views <- views[,c("article", "project", "timestamp", "views")]
  21 views$timestamp <- fct_explicit_na(as.character(views$timestamp))
  22
  23
  24 ### Sorts and groups at the same time
  25 views.by.proj.date <- arrange(group_by(views, project, timestamp),
  26                         desc(views))
  27
  28 ### Export just the top 10 by pageviews
  29 write.table(head(views.by.proj.date, 10),
  30             file="output/top10_views_by_project_date.csv", sep=",",
  31             row.names=FALSE)
  32
  33 ### A simple visualization
  34 p <- ggplot(data=views.by.proj.date, aes(views))
  35
  36 ## Density plot with log-transformed axis
  37 p + geom_density() + scale_x_log10(labels=comma)
  38
  39
  40