July 2017
Beginner to intermediate
312 pages
7h 27m
English
We use a similar approach for sentiment to keep the results comparable:
dx = df[df.index > '2013-11-18']ax = dx.groupby(dx.index.weekday)['sentiment'].mean().plot(kind = 'bar', title = 'Average sentiment by day of week')ax.set_xticklabels(['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'])plt.show()

The analysis by the day of the week shows an interesting trend. While the maximum number of comments is on Friday, the most positive sentiment is on Thursday. Based on these two extreme days, we notice a correlation between the number of comments and its sentiment saying that the lesser number ...
Read now
Unlock full access