如何为每个组生成摘要统计信息并进行转置？

Question

我有两个如下所示的表

我正在尝试类似以下的内容

select person_id,
   min(value),max(value),count(value),stddev(value)
   percentile_cont(0.25) within group (order by value_as_number asc) as "25pc",
   percentile_cont(0.75) within group (order by value_as_number asc) as "75pc"
from cdm.measurement
group by person_id

如果是person doesn't have a specific reading, the value should NA as shown in screenshot below

我想做两件事

1）创建与Readings表中的读数数一样多的列（只有唯一的ID）。 例如，如果“读数”表具有800个阅读ID，则我们将为一个人提供800 * 6 = 4800列。 6 is used here because of min,max,count,stddev,25th percentile,75 percentile。列的名称将为R_name followed by 25%,75%,min,max etc。例如：Read_1_25％，Read_1_min

2）如果某人没有任何读数，则其缺少读数的值为NA。例如，“ person_id = 1”仅具有R1读数。因此其余4794（4800-6）列将为NA

我希望我的输出如下所示。 由于图像较宽，请单击图像将其放大。忽略我的屏幕截图中值的正确性。格式是我在寻求您的帮助

Answer 1

如果我理解正确，则可以使用条件聚合：

select person_id,
       count(*) filter (where reading = 'R_1') as cnt_r_1,
       min(value) filter (where reading = 'R_1') as min_r_1,
       max(value) filter (where reading = 'R_1') as max_r_1,
       avg(value) filter (where reading = 'R_1') as avg_r_1,
       stdev(value) filter (where reading = 'R_1') as stdev_r_1,
       count(*) filter (where reading = 'R_2') as cnt_r_2,
       min(value) filter (where reading = 'R_2') as min_r_2,
       max(value) filter (where reading = 'R_2') as max_r_2,
       avg(value) filter (where reading = 'R_2') as avg_r_2,
       stdev(value) filter (where reading = 'R_2') as stdev_r_2,
       . . .
from t
group by person_id;