Was the NYT the end of Wordle's popularity? A story told with tweet sentiment
Choosing our Twitter query
print(WORDLE_QUERY)
Tweet counts over time
Analyzing sentiment to find out how people feel about Wordle
positive_negative_sentiments_df = mean_sentiment_df[
(mean_sentiment_df["label"] == "Positive")
| (mean_sentiment_df["label"] == "Negative")
]
positive_negative_sentiments_df = positive_negative_sentiments_df.merge(
tweet_counts_df, on="day"
)
positive_negative_sentiments_df["tweet_count_estimated"] = (
positive_negative_sentiments_df["mean_sentiment"]
* positive_negative_sentiments_df["count"]
)
fig = px.line(
positive_negative_sentiments_df,
x="day",
y="tweet_count_estimated",
color="label",
color_discrete_map=plot_colors,
labels={
"tweet_count_estimated": "Estimated tweet count",
"day": "Day",
"label": "Label",
},
)
fig.update_layout(
legend=dict(orientation="h", yanchor="bottom", y=1.02, xanchor="right", x=1)
)
fig
Understanding the sentiment timeline
December 17: The honeymoon period
December 27: What happened over Christmas?
February 1: Peak Wordle
February 17: All-time low
March 21: Recovering positivity
Are more tweets bad news for Wordle?
Conclusion
Choose a day yourself
summarize_date
daily_summary = summarize_day_of_interest(summarize_date)
daily_summary["show_summary"]()
Top positive tweets
daily_summary["positive_tweets_df"]
Top negative tweets (warning: explicit language)
daily_summary["negative_tweets_df"]