当前位置:   article > 正文

HiveSQL经典面试题(建议点赞收藏)_hive sql 面试题

hive sql 面试题

目录

经典面试系列

每科成绩都大于80分的学生信息

连续登录问题

行列转换问题

留存问题:看当天登录后第N天是否登录

TopN问题

累计计算问题

HSQL进阶版

 直播间在线最大人数

SQL循环

计算中位数

产生连续数值


经典面试系列

  1. 每科成绩都大于80分的学生信息

    1. 建表+初始化
      1. --创建表scdn_student_score_test 并且初始化三个学生成绩
      2. create table hdw_tmp_dev.scdn_student_score_test as
      3. select '张三' as name,'数学' as subject ,'80' as score
      4. union all
      5. select '张三' as name,'语文' as subject ,'90' as score
      6. union all
      7. select '张三' as name,'英语' as subject ,'90' as score
      8. union all
      9. select '李四' as name,'数学' as subject ,'90' as score
      10. union all
      11. select '李四' as name,'语文' as subject ,'90' as score
      12. union all
      13. select '李四' as name,'英语' as subject ,'70' as score
      14. union all
      15. select '王五' as name,'数学' as subject ,'90' as score
      16. union all
      17. select '王五' as name,'语文' as subject ,'90' as score
      18. union all
      19. select '王五' as name,'英语' as subject ,'50' as score
      20. --查询结果显示
      21. | scdn_student_score_test.name | scdn_student_score_test.subject | scdn_student_score_test.score |
      22. +-------------------------------+----------------------------------+--------------------------------+
      23. | 张三 | 数学 | 80 |
      24. | 张三 | 语文 | 90 |
      25. | 张三 | 英语 | 90 |
      26. | 李四 | 数学 | 90 |
      27. | 李四 | 语文 | 90 |
      28. | 李四 | 英语 | 70 |
      29. | 王五 | 数学 | 90 |
      30. | 王五 | 语文 | 90 |
      31. | 王五 | 英语 | 50 |
      32. +-------------------------------+----------------------------------+---------------------
      33. --求所有学科成绩都大于等于80分的学生姓名
    2. 思路一(思维转换):所有问题找最小(最大)
      1. select
      2. t1.name
      3. ,t1.min_score
      4. from
      5. (
      6. select
      7. name
      8. ,min(score) as min_score
      9. from hdw_tmp_dev.scdn_student_score_test
      10. group by name
      11. ) as t1 --求出最小的成绩
      12. where t1.min_score >=80 --最小成绩大于等于80,则这个学生的所有成绩都会大于等于80
    3. 执行结果
    4. 思路二(巧用左关联进行筛选):
      1. select
      2. t1.name
      3. from
      4. (
      5. select
      6. name
      7. from hdw_tmp_dev.scdn_student_score_test
      8. group by name
      9. ) as t1
      10. left join
      11. (
      12. select
      13. name
      14. from hdw_tmp_dev.scdn_student_score_test
      15. where score <80
      16. group by name
      17. ) as t2 on t1.name = t2.name
      18. where t2.name is null
  2. 连续登录问题

    1. 建表+初始化
      1. create table hdw_tmp_dev.csdn_user_login_test as
      2. select 'xiaoming' as user_name,'2024-01-01' as login_date
      3. union all
      4. select 'xiaoming' as user_name,'2024-01-02' as login_date
      5. union all
      6. select 'xiaoming' as user_name,'2024-01-03' as login_date
      7. union all
      8. select 'xiaoming' as user_name,'2024-01-04' as login_date
      9. union all
      10. select 'xiaoming' as user_name,'2024-01-05' as login_date
      11. union all
      12. select 'dahuang' as user_name,'2024-01-02' as login_date
      13. union all
      14. select 'dahuang' as user_name,'2024-01-03' as login_date
      15. union all
      16. select 'dahuang' as user_name,'2024-01-04' as login_date
      17. union all
      18. select 'dahuang' as user_name,'2024-01-05' as login_date
      19. union all
      20. select 'lucky_dog' as user_name,'2024-01-01' as login_date
      21. union all
      22. select 'lucky_dog' as user_name,'2024-01-03' as login_date
      23. union all
      24. select 'lucky_dog' as user_name,'2024-01-04' as login_date
      25. union all
      26. select 'lucky_dog' as user_name,'2024-01-05' as login_date
    2. Sql参考
      1. select
      2. t2.user_name
      3. ,t2.date_begin_flag
      4. ,count(1) as max_login_days
      5. from
      6. (
      7. select
      8. t1.user_name
      9. ,t1.login_date
      10. ,date_sub(date(login_date),t1.rn) as date_begin_flag
      11. from
      12. (
      13. select
      14. user_name
      15. ,login_date
      16. ,row_number()over(partition by user_name order by login_date) as rn
      17. from hdw_tmp_dev.csdn_user_login_test
      18. ) as t1
      19. ) as t2
      20. group by t2.user_name
      21. ,t2.date_begin_flag
    3. 执行过程+结果
  3. 行列转换问题

    1. 建表+初始化:参考问题一的表
      1. | t1.name | t1.subject | t1.score |
      2. +-------------------------------+----------------------------------+-----------------+
      3. | 张三 | 数学 | 80 |
      4. | 张三 | 语文 | 90 |
      5. | 张三 | 英语 | 90 |
      6. | 李四 | 数学 | 90 |
      7. | 李四 | 语文 | 90 |
      8. | 李四 | 英语 | 70 |
      9. | 王五 | 数学 | 90 |
      10. | 王五 | 语文 | 90 |
      11. | 王五 | 英语 | 50 |
      12. +-------------------------------+----------------------------------+------------------
    2. 行专列Sql参考
      1. select
      2. name
      3. ,max(case when subject = '数学' then score end) as math_score
      4. ,max(case when subject = '语文' then score end) as china_score
      5. ,max(case when subject = '英语' then score end) as english_score
      6. from hdw_tmp_dev.scdn_student_score_test
      7. group by name
    3. 行专列执行结果
    4. 列转行:可以参考建表初始化语句(union all)
      1. select
      2. name
      3. ,collect_set(subject) as subject_set
      4. from hdw_tmp_dev.scdn_student_score_test
      5. group by name
    5. 每个学生选课结果(多行变一行):

    6.  将上面的结果展开(一行变多行)数据准备:

      1. create table hdw_tmp_dev.scdn_student_score_test_collect as
      2. select 'zhangsan' as name ,'"数学","语文","英语"' as subject_list
      3. union all
      4. select 'lisi' as name ,'"美术","生物","物理"' as subject_list
      5. union all
      6. select 'wangwu' as name ,'"计算机","日语","韩语"' as subject_list
    7. 炸开代码
      1. select
      2. name
      3. ,subject_list
      4. ,subject_name
      5. from hdw_tmp_dev.scdn_student_score_test_collect
      6. lateral view explode(split(subject_list,',')) extend_sub as subject_name
    8. 结果

  4. 留存问题:看当天登录后第N天是否登录

    1. 建表+初始化:参照连续登录表
    2. Sql参考
      1. --方案一:利用lead(日期,N)是否等于 当天登录实践+N天
      2. select
      3. t1.user_name
      4. ,t1.logon_date
      5. ,case when lead1_logon_date = date_add(logon_date,1) then '1天留存' end as 1day_remain
      6. ,case when lead3_logon_date = date_add(logon_date,3) then '3天留存' end as 3day_remain
      7. from
      8. (
      9. select
      10. user_name
      11. ,logon_date
      12. ,lead(user_name,1)over(partition by user_name order by logon_date) as lead1_user_name
      13. ,lead(logon_date,1)over(partition by user_name order by logon_date) as lead1_logon_date
      14. ,lead(user_name,3)over(partition by user_name order by logon_date) as lead3_user_name
      15. ,lead(logon_date,3)over(partition by user_name order by logon_date) as lead3_logon_date
      16. from hdw_tmp_dev.csdn_user_logon_test
      17. ) as t1
      18. --方案二:
      19. select
      20. t2.first_log_date as first_log_date
      21. ,count(t2.user_id) as new_user_cnt --新用户数
      22. ,count(t3.user_id) as next_user_id --次日回访用户数
      23. ,count(t4.user_id) as 30_user_id --30天回访用户数
      24. ,count(t3.user_id)/count(t2.user_id) as next_back_rate --次日回访率
      25. ,count(t4.user_id)/count(t2.user_id) as 30_back_rate --30天回访率
      26. from
      27. (
      28. select
      29. first_log_date
      30. ,user_id
      31. ,date_add(first_log_date,1) as next_log_date
      32. ,date_add(first_log_date,29) as 30_log_date
      33. from
      34. (
      35. select
      36. user_id
      37. ,log_time
      38. ,first_value(date(log_time))over(partition by user_id) as first_log_date
      39. from user_log
      40. ) as t1
      41. group by first_log_date
      42. ,user_id
      43. ) as t2
      44. left join
      45. (
      46. select
      47. user_id
      48. ,date(log_date) as log_date
      49. from user_log
      50. group by user_id
      51. ,date(log_date) as log_date
      52. ) as t3 on t2.user_id = t2.user_id and t2.next_log_date = t3.log_date
      53. left join
      54. (
      55. select
      56. user_id
      57. ,date(log_date) as log_date
      58. from user_log
      59. group by user_id
      60. ,date(log_date) as log_date
      61. ) as t4 on t2.user_id = t4.user_id and t2.30_log_date = t4.log_date
      62. group by t2.first_log_date
    3. 方案一执行结果
  5. TopN问题

    1. 建表+初始化
      1. create table hdw_tmp_dev.scdn_student_score_test1 as
      2. select '张三' as name,'数学' as subject ,'80' as score
      3. union all
      4. select '张三' as name,'语文' as subject ,'90' as score
      5. union all
      6. select '张三' as name,'英语' as subject ,'90' as score
      7. union all
      8. select '李四' as name,'数学' as subject ,'90' as score
      9. union all
      10. select '李四' as name,'语文' as subject ,'90' as score
      11. union all
      12. select '李四' as name,'英语' as subject ,'70' as score
      13. union all
      14. select '王五' as name,'数学' as subject ,'90' as score
      15. union all
      16. select '王五' as name,'语文' as subject ,'90' as score
      17. union all
      18. select '王五' as name,'英语' as subject ,'50' as score
      19. union all
      20. select '小明' as name,'数学' as subject ,'88' as score
      21. union all
      22. select '小明' as name,'语文' as subject ,'99' as score
      23. union all
      24. select '小明' as name,'英语' as subject ,'77' as score
      25. union all
      26. select '小文' as name,'数学' as subject ,'66' as score
      27. union all
      28. select '小文' as name,'语文' as subject ,'89' as score
      29. union all
      30. select '小文' as name,'英语' as subject ,'90' as score
    2. Sql参考:求每科前三名对应的人员的成绩单
      1. select
      2. *
      3. from
      4. (
      5. select
      6. name
      7. ,subject
      8. ,score
      9. ,row_number()over(partition by subject order by score desc) as rn
      10. from hdw_tmp_dev.scdn_student_score_test1
      11. ) as t1
      12. where rn<=3
    3. 执行结果
  6. 累计计算问题

    1. 建表+初始化
      1. create table hdw_tmp_dev.user_sale_date as
      2. select '001' as user_id,'2024-02-01' as sale_date, 100 as amount
      3. union all
      4. select '001' as user_id,'2024-02-02' as sale_date, 200 as amount
      5. union all
      6. select '001' as user_id,'2024-02-03' as sale_date, 300 as amount
      7. union all
      8. select '001' as user_id,'2024-02-04' as sale_date, 400 as amount
      9. union all
      10. select '001' as user_id,'2024-02-05' as sale_date, 500 as amount
      11. union all
      12. select '001' as user_id,'2024-02-06' as sale_date, 600 as amount
    2. SQL逻辑
      1. select
      2. user_id
      3. ,sale_date
      4. ,amount
      5. ,sum(amount)over(partition by user_id order by sale_date) as accuma_amount --按日期逐渐累加
      6. ,sum(amount)over(partition by user_id) as total_amount --按人汇总
      7. ,avg(amount)over(partition by user_id) as avg_amount --安人平均每天
      8. ,max(amount)over(partition by user_id) as max_amount --单日最大销售
      9. ,min(amount)over(partition by user_id) as min_amount --单日最小销售
      10. from hdw_tmp_dev.user_sale_date
    3. 结果展示

HSQL进阶版

  1.  直播间在线最大人数

    1. 建表+初始化
      1. create table hdw_tmp_dev.csdn_user_login_time_detail as
      2. select '001' as user_id,'2024-02-01 10:00:00' as begin_time,'2024-02-01 12:00:00' as end_time
      3. union all
      4. select '002' as user_id,'2024-02-01 11:00:00' as begin_time,'2024-02-01 13:00:00' as end_time
      5. union all
      6. select '003' as user_id,'2024-02-01 11:00:00' as begin_time,'2024-02-01 14:00:00' as end_time
      7. union all
      8. select '004' as user_id,'2024-02-01 15:00:00' as begin_time,'2024-02-01 16:00:00' as end_time
    2. SQL逻辑
      1. select
      2. t1.user_id
      3. ,t1.time1
      4. ,t1.flag
      5. ,sum(flag)over(order by t1.time1) as user_cnt
      6. from
      7. (
      8. select
      9. user_id
      10. ,begin_time as time1
      11. ,1 as flag
      12. from hdw_tmp_dev.csdn_user_login_time_detail
      13. union all
      14. select
      15. user_id
      16. ,end_time
      17. ,-1 as flag
      18. from hdw_tmp_dev.csdn_user_login_time_detail
      19. ) as t1
    3. 结果展示
  2. SQL循环

    1. 建表+初始化
      1. create table hdw_tmp_dev.cycle_1 as
      2. select '1011' as a
      3. union all
      4. select '0101' as a
    2. SQL逻辑
      1. select
      2. a,
      3. concat_ws(",",collect_list(cast(index as string))) as res
      4. from (
      5. select
      6. a,
      7. index+1 as index,
      8. chr
      9. from (
      10. select
      11. a,
      12. concat_ws(",",substr(a,1,1),substr(a,2,1),substr(a,3,1),substr(a,-1)) str
      13. from hdw_tmp_dev.cycle_1 as t8
      14. ) tmp1
      15. lateral view posexplode(split(str,",")) t as index,chr
      16. where chr = "1"
      17. ) tmp2
      18. group by a;
    3. 结果展示
  3. 计算中位数

    1. 建表+初始化:
      1. create table hdw_tmp_dev.user_sale_date as
      2. select '001' as user_id,'2024-02-01' as sale_date, 100 as amount
      3. union all
      4. select '001' as user_id,'2024-02-02' as sale_date, 200 as amount
      5. union all
      6. select '001' as user_id,'2024-02-03' as sale_date, 300 as amount
      7. union all
      8. select '001' as user_id,'2024-02-04' as sale_date, 400 as amount
      9. union all
      10. select '001' as user_id,'2024-02-05' as sale_date, 500 as amount
      11. union all
      12. select '001' as user_id,'2024-02-06' as sale_date, 600 as amount
    2. SQL逻辑
      1. select
      2. t1.user_id
      3. ,t1.sale_date
      4. ,t1.amount
      5. ,avg(t1.amount)over(partition by t1.user_id) as zhongwenshu
      6. from
      7. (
      8. select
      9. user_id
      10. ,sale_date
      11. ,amount
      12. ,row_number()over(partition by user_id order by amount) as rn
      13. from hdw_tmp_dev.user_sale_date
      14. ) as t1
      15. left join
      16. (
      17. select
      18. user_id
      19. ,count(1) as cnt
      20. from hdw_tmp_dev.user_sale_date
      21. group by user_id
      22. ) as t2 on t1.user_id = t2.user_id
      23. where t1.rn in (cnt/2,(cnt+1)/2,cnt/2+1)
      24. --总个数为奇数命中(cnt+1)/2;总个数为偶数命中:cnt/2,cnt/2+1,两数相加求平均值
    3. 结果展示
  4. 产生连续数值

    1. SQL逻辑
      1. --产生1到10的连续数据
      2. select
      3. start_id + pos as id
      4. ,pos
      5. ,val
      6. from(
      7. select
      8. 1 as start_id,
      9. 10 as end_id
      10. ) m lateral view posexplode(split(space(end_id - start_id),'')) t as pos, val
      11. --方案二
      12. select
      13. row_number() over() as id
      14. from
      15. (select split(space(99), '') as x) t
      16. lateral view
      17. explode(x) ex;
    2. 结果展示
声明:本文内容由网友自发贡献,版权归原作者所有,本站不承担相应法律责任。如您发现有侵权的内容,请联系我们。转载请注明出处:【wpsshop博客】
推荐阅读
相关标签
  

闽ICP备14008679号