What percentage of tweets have a valid negative reason recorded after data preprocessing?
What percentage of tweets have a valid negative reason recorded after data preprocessing?: a task in SmolDataEnvs-harbor-train-sandbox: 📊 SmolDataEnvs: Harbor (train) (Harbor dataset). Files (in /home/user/input, no subfolders): - Tweets.csv - database.sqlite
The task
Files (in /home/user/input, no subfolders): - Tweets.csv - database.sqlite