0

Please login or register to vote for this query.

(click on this box to dismiss)

Q&A for people interested in the history and origins of science and mathematics

WITH main AS (SELECT post_year,
       post_month,
       tag,
       month_count -- optional, not needed in main table
       
FROM (

  SELECT datepart(year,Posts.CreationDate) as post_year, 
         datepart(month,Posts.CreationDate) as post_month, 
         tags.TagName as tag, 
         COUNT(*) as month_count
         
  FROM PostTags as pt
    INNER JOIN Posts
    ON pt.PostId = Posts.Id
    INNER JOIN Tags as tags 
    ON pt.TagId = tags.Id
    
  WHERE tags.TagName in ('tensorflow', 'pytorch', 'scipy',
                         'keras', 'scikit-learn', 'numpy', 'pandas')
  AND Posts.PostTypeId = 1
  GROUP BY datepart(year,Posts.CreationDate), 
           datepart(month,Posts.CreationDate), 
           tags.TagName
           
     ) AS subquery
     
     
GROUP BY post_year, 
         post_month, 
         tag,
         month_count), 

sub AS (SELECT datepart(year,Posts.CreationDate) as post_year, 
               datepart(month,Posts.CreationDate) as post_month,
               count(*) as total
        FROM Posts WHERE PostTypeId = 1
        GROUP BY datepart(year,Posts.CreationDate), 
                 datepart(month,Posts.CreationDate))
              

SELECT main.post_year as 'year',
       main.post_month as 'month',
       tag as 'tag',
       month_count as 'month count per tag',
       total as 'overall count',
       month_count * 100.0 / total as 'percent',
       SUM(month_count) OVER(PARTITION BY tag
                             ORDER BY main.post_year,
                                      main.post_month
                             ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) as 'running count'
FROM main INNER JOIN sub ON main.post_year = sub.post_year 
                         WHERE main.post_month = sub.post_month
ORDER BY main.post_year,
         main.post_month,
         tag;

fork query

edited may 21

created may 18 anonymous