3 ms·
I had a hacky implementation of this in a previous analysis: http://minimaxir.com/img/hn-top-submitters.png http://minimaxir.com/img/hn-top-submitters.png I ca
by minimaxir 11y ago
I had a hacky implementation of this in a previous analysis: http://minimaxir.com/img/hn-top-submitters.png http://minimaxir.com/img/hn-top-submitters.png
I can sorta recreate it with this BigQuery:
SELECT author, COUNT(author) as num_submissions,
ROUND(SUM(score >= 10)/COUNT(author), 3) AS perc
FROM [fh-bigquery:hackernews.stories]
GROUP BY author
HAVING num_submissions >= 10
ORDER BY perc DESC
LIMIT 100
Which results in this output: https://docs.google.com/spreadsheets/d/1EHqYvvAh9zUSUSrn_6t96ENtXrMRKB-ntc57TViBZOE/edit?usp=sharing https://docs.google.com/spreadsheets/d/1EHqYvvAh9zUSUSrn_6t9...
- gus_massa 11y agoFor the visualization, I prefer a dispersion graph like this: https://imgur.com/0k9loNC https://imgur.com/0k9loNC In case you are wondering, the four outliers at the right are: author num_sub perc pkrumins 133 75.2% whoishiring 126 100.0% jaf12duke 123 85.4% katm 82 85.4%