计算MySQL中其他平均值的平均值

Question

I have this table where I want to calculate the average response time from all users based on the cat_id and from a join based on the user_id 's country which are stored in another table.

What is the proper way to find the average time based on the user's country?

+---------+--------+-------+-------+-------+-------+-------+-------+-------+
| user_id | cat_id | time1 | time2 | time3 | time4 | time5 | time6 | time7 |
+---------+--------+-------+-------+-------+-------+-------+-------+-------+
|      11 |      1 | 10000 |  3500 | 10000 | 10000 |  7000 |  6000 |  6500 |
|       0 |      1 |  6139 |  2640 |  5438 |  8136 |  7157 |  8723 |  4677 |
|      11 |      3 |  7500 |  5500 | 10000 |  5500 |  7500 |  4000 |  9500 |
|       0 |      3 |  7462 |  8907 |  7513 |  4181 |  7275 |  2074 |  4444 |
|       1 |      1 |  8500 |  2000 |  8500 |  1500 | 10000 |  6500 | 10000 |
|       0 |      1 |  8885 |  7240 |  2153 |  6115 |  6330 |  6151 |  4325 |
|       1 |      1 | 10000 |  3000 | 10000 |  5500 | 10000 |  9000 | 10000 |
|       0 |      1 |  6188 |  3280 |  4725 |  5662 |  3344 |  8792 |  7560 |
|       1 |      4 |  4000 |  4500 | 10000 |  5000 |  3000 |  1500 |  2112 |
|       0 |      4 |  2750 |  4861 |  3413 |  4174 |  3080 |  3213 |  7932 |
+---------+--------+-------+-------+-------+-------+-------+-------+-------+

I've tried: (AVG(time1 + time 2 + ... + time7) / 7) but when I cross-check the results are not correct.

Thanks in advance for any help :)

Answer 1

First of all there's a problem with your parenthesis, which should be

AVG((time1 + time2 + ... + time7) / 7)

Then you may need to add something in your group by clause, depending on the level you want the average to be calculated at.

Answer 2

首先，您应该计算行(time1 + time2 + ... time7)/7平均值，然后计算所有行的AVG((time1 + time2 + ... time7)/7)

Answer 3

I think your problem is taking an average of averages. I would first calculate the sum and count for each row and left join to add the country. Once you have your tabular data, you can summarize and do averages across country, cat, or both. Something like this:

select country, cat_id, (sum(total_time)/sum(count)) as avg_response_time
from (
select a.user_id
      ,a.cat_id
      ,b.country
      ,(a.time1 + a.time2 + a.time3 + a.time4 + a.time5 + a.time6 + a.time7) as total_time
      ,7 as count
from a
left outer join b
on a.user_id = b.user_id
)
group by country, cat_id