pointing at updated data url, adding explicit NA handling to factor, cutting unnecess...

[covid19.git] / wikipedia_views / analysis / pageview_example.R

diff --git a/wikipedia_views/analysis/pageview_example.R b/wikipedia_views/analysis/pageview_example.R
index 8a7aba35dfe6f5827a7a1debc0192730658e3214..fb5359aa64c856bec978b1a5bcfc9712de82c907 100644 (file)
--- a/wikipedia_views/analysis/pageview_example.R
+++ b/wikipedia_views/analysis/pageview_example.R
@@ -4,13 +4,12 @@
 ### Minimal example analysis file using pageview data
 
 library(tidyverse)
-library(ggplot2)
 library(scales)
 
-### Import and cleanup data
+### Import and cleanup one datafile from the observatory
 
 DataURL <-
-    url("https://github.com/CommunityDataScienceCollective/COVID-19_Digital_Observatory/raw/master/wikipedia_views/data/dailyviews2020032600.tsv")
+    url("https://covid19.communitydata.science/datasets/wikipedia/digobs_covid19-wikipedia-enwiki_dailyviews-20200401.tsv")
 
 views <-
     read.table(DataURL, sep="\t", header=TRUE, stringsAsFactors=FALSE) 
@@ -30,12 +29,14 @@ views <-
 ### (see https://www.tidyverse.org for more info)
 
 views <- views[,c("article", "project", "timestamp", "views")]
-views$timestamp <- factor(views$timestamp)
+views$timestamp <- fct_explicit_na(views$timestamp)
+
 
 ### Sorts and groups at the same time
 views.by.proj.date <- arrange(group_by(views, project, timestamp),
                         desc(views))
 
+
 ### Export just the top 10 by pageviews
 write.table(head(views.by.proj.date, 10),
             file="output/top10_views_by_project_date.csv", sep=",",