MySQL的group_concat 函数经常实用，功能蛮强大的。

MySQL创建表：

create table kpi(emp_no varchar(8),performance varchar(32),month varchar(32));

clickhouse 创建表：
create table kpi(emp_no varchar(8),performance varchar(32),month varchar(32)) engine=Memory;


分别在MySQL和Clickhouse中插入数据：
insert into kpi values (10,'A','2020-01'),(10,'A','2020-02'),(10,'C','2020-03'),(10,'B','2020-04'),(10,'A','2020-05'),(10,'A','2020-06');
insert into kpi values (20,'A','2020-01'),(20,'B','2020-02'),(20,'C','2020-03'),(20,'C','2020-04'),(20,'A','2020-05'),(20,'D','2020-06');
insert into kpi values (30,'C','2020-03'),(30,'C','2020-04'),(30,'B','2020-05'),(30,'B','2020-06');

数据查询如下：

MySQL的group_concat 功能展示：

mysql> select emp_no,group_concat(performance order by month separator '-') kpi_list,group_concat(distinct performance order by month separator '-') kpi_uniq,group_concat(distinct performance order by month desc separator '-') kpi_uniq_desc from kpi group by emp_no;  
+--------+-------------+----------+---------------+
| emp_no | kpi_list    | kpi_uniq | kpi_uniq_desc |
+--------+-------------+----------+---------------+
| 10     | A-A-C-B-A-A | A-C-B    | B-C-A         |
| 20     | A-B-C-C-A-D | A-B-C-D  | D-C-B-A       |
| 30     | C-C-B-B     | C-B      | B-C           |
+--------+-------------+----------+---------------+
3 rows in set (0.00 sec)

kpi_list:按照月份依次显示每个月的绩效
kpi_uniq:上半年获得的绩效 等级（绩效去重）
kpi_uniq_desc ：去重后的绩效反向排序

 Clickhouse的Group_concat 实现：

SELECT 
    emp_no,
    groupArray(performance) AS kpi_asc,
    arrayStringConcat(kpi_asc, '-') AS kpi_list,
    arrayReverse(kpi_asc) AS kpi_desc,
    groupUniqArray(performance) AS kpis,
    arraySort(kpis) AS kpi_uniq,
    countEqual(kpi_asc, 'A') AS A_cnt,
    countEqual(kpi_asc, 'B') AS B_cnt,
    countEqual(kpi_asc, 'C') AS C_cnt,
    countEqual(kpi_asc, 'D') AS D_cnt
FROM kpi
GROUP BY emp_no
ORDER BY emp_no ASC

┌─emp_no─┬─kpi_asc───────────────────┬─kpi_list────┬─kpi_desc──────────────────┬─kpis──────────────┬─kpi_uniq──────────┬─A_cnt─┬─B_cnt─┬─C_cnt─┬─D_cnt─┐
│ 10     │ ['A','A','C','B','A','A'] │ A-A-C-B-A-A │ ['A','A','B','C','A','A'] │ ['B','A','C']     │ ['A','B','C']     │     4 │     1 │     1 │     0 │
│ 20     │ ['A','B','C','C','A','D'] │ A-B-C-C-A-D │ ['D','A','C','C','B','A'] │ ['B','A','D','C'] │ ['A','B','C','D'] │     2 │     1 │     2 │     1 │
│ 30     │ ['C','C','B','B']         │ C-C-B-B     │ ['B','B','C','C']         │ ['B','C']         │ ['B','C']         │     0 │     2 │     2 │     0 │
└────────┴───────────────────────────┴─────────────┴───────────────────────────┴───────────────────┴───────────────────┴───────┴───────┴───────┴───────┘

3 rows in set. Elapsed: 0.007 sec. 

结论:要等同于MySQL中的group_concat 功能需要使用到Clickhouse中的几个函数组合使用：

groupArray 行专列

groupUniqArray  等同于mysql中的 group_concat(distinct ..）

arrayStringConcat  等同于group_concat 子语句中的separator '-' 语句

countEqual 则统计数组中元素出现的次数，MySQL实现成功要麻烦点。