@@ -38,7 +38,7 @@ As the old saying goes, every analysis needs its data. While I generally prefer
And while `ggplot2` - especially when combined with `ggthemes` - comes with beautiful themes by default, I most often cannot resist to create a custom theme for each project. With the colour scheme below, I tried to match the visual appearance of Shirin's follower analysis. And yes, it would have been easier, albeit less fun, if I had followed her example and simply used `tidyquant`.
xlab("") + ylab("") + ggtitle("Top locations of Open Science Twitter MOOC-ers", subtitle="") +
xlab("") + ylab("Count") + ggtitle("Top locations of Open Science Twitter MOOC-ers", subtitle="") +
coord_flip() +viz_theme
```
@@ -214,11 +214,11 @@ ggplot() +
In a slightly more advanced version of this map, I added further information on the follower count of the geocoded users, which results in a similar picture: The majority of the influential Open Science Twitter MOOC-ers are based in Western Europe and the United States. I will revisit this notion later on in the blog post when conducting a more fine-grained analysis of the most prominent followers.
ggtitle(label="Count of Open Science Twitter MOOC-ers by gender", subtitle="Gender predicted with genderize.io (random sample of N = 1000 followers)") +
ggtitle(label="Open Science Twitter MOOC-ers by gender", subtitle="Gender predicted with genderize.io (random sample of N = 1000 followers)") +
ggtitle(label='Number of Open Science Twitter MOOC-ers by account status', subtitle="Influencer: at least 500 followers and at least thrice as many followers than friends \nVerified: official verification status") + ylim(0, 6000) +
ggtitle(label='Open Science Twitter MOOC-ers by account status', subtitle="Influencer = at least 500 followers and at least thrice as many followers than friends \nVerified = official verification status") + ylim(0, 6000) +
scale_color_continuous("Number of days since \naccount was created", low="#91aac3", high="#2C3E50") +
xlab("Number of followers (log2)") + ylab("Average number of tweets per day (log2)") + ggtitle("Correlation between the Open Science MOOC-ers followers count and tweets per day", subtitle="") +
xlab("Number of followers (log2)") + ylab("Average number of tweets per day (log2)") + ggtitle("Correlation between followers count and tweets per day", subtitle="") +
xlab("") + ylab("") + ggtitle("Count of positive and negative sentiments in the Open Science MOOC-ers profile descriptions", subtitle="Sentiment lexicon by Bing Liu and collaborators") + ylim(0, 2000) +
xlab("") + ylab("Count") + ggtitle("Positive and negative sentiments in Open Science MOOC-ers' profiles", subtitle="Sentiment lexicon by Bing Liu and collaborators") + ylim(0, 2000) +
xlab("Sentiment") + ylab("Density") + ggtitle("Distribution of sentiments in the Open Science MOOC-ers profile descriptions", subtitle="Sentiment lexicon by Bing Liu and collaborators") + ylim(0, 1.6) +
xlab("Sentiment") + ylab("Density") + ggtitle("Distribution of sentiments in Open Science MOOC-ers' profiles", subtitle="Sentiment lexicon by Bing Liu and collaborators") + ylim(0, 1.6) +
xlab("") + ylab("") + ggtitle("Most common positive and negative words in the Open Science MOOC-ers profile descriptions", subtitle="Sentiment lexicon by Bing Liu and collaborators") +
xlab("") + ylab("") + ggtitle("Most common positive and negative words in Open Science MOOC-ers' profiles", subtitle="Sentiment lexicon by Bing Liu and collaborators") +
0 comments on commit
84d8b6a