我也在上同样的Coursera课程，这就是我的解决方案。请注意，在其他SQL语言中，Covar和COR函数使这一过程变得容易得多。无法计算R函数，因为SQLite中没有SQRT()函数。

select avg( (review_count - avg_x) * (fans - avg_y) )*avg( (review_count - avg_x) * (fans - avg_y) )/(var_x*var_y) as R2 
from user, (select 
      avg_x,
      avg_y,
      avg((review_count - avg_x)*(review_count - avg_x)) as var_x, 
      avg((fans - avg_y)*(fans - avg_y)) as var_y 
      from user, (select 
          avg(review_count) as avg_x, 
          avg(fans) as avg_y 
          from user)
  );

SQLite没有内置的平方根函数。也不是内置的电源功能。但如果你能满足于R^2，那么我认为公式是：

select (sum( (x - avg_x) * (y - avg_y) ) * sum( (x - avg_x) * (y - avg_y) ) /
       nullif(sum( (x - avg_x) * (x - avg_x) ) * 
              sum( (y - avg_y) * (y - avg_y)), 0
             ) as r_square
from (select t.*, review_cnt as x, fans as y,
             avg(review_cnt) over () as avg_x,
             avg(fans) over () as avg_y
      from t
     ) t;

您可以对照Wikipedia page进行检查。