在线教育仓库示例-学生出勤看板
总结需求:
第一类: 指标: 计算 出勤人数, 出勤率, 迟到人数, 迟到率
涉及维度:
时间维度: 年 月 天 上午 下午 晚自习
班级维度:
学生维度:
涉及表:
course_table_upload_detail: 日志课程明细表 (课表) (维度表)
tbh_student_signin_record: 学生打卡记录表 (事实表)
tbh_class_time_table: 班级作息时间表 (维度表)
关联条件:
学生打卡表.class_id = 课程表.class_id
班级作息时间表.id = 学生打卡表.time_table_id
涉及到字段:
时间维度: 课程表.class_date
班级维度: 课程表.class_id
学生维度: 学生打卡表.student_id
指标字段: 学生打卡表.signin_time(打卡时间)
计算方案:
先判断是否出勤
情况1: 出勤了, 再次判断是否是正常出勤和迟到出勤
情况2: 未出勤, 认为没有来
指标判断指标:
作息时间表:
morning_begin_time
morning_end_time
afternoon_begin_time
afternoon_end_time
evening_begin_time
evening_end_time
过滤操作
1) ifnull(ctud.content,'') != '' 相当于 ctud.content != null and ctud.content != ''
转换为hive的操作
nvl(ctud.content,'') != ''
2) 将content中为开班典礼数据过滤掉
ctud.content != '开班典礼'
3) 确保打卡表中学生数据都是开启公屏
学生打卡表.share_state=1
判断学生出勤状态: 0(正常出勤) 1(迟到出勤) 2(缺勤
首先根据打卡时间, 如果打卡的时间在 上课的前40分钟内 ~ 上课截止时间内
情况1: 出勤了, 接着判断 是否是正常出勤还是迟到出勤,
如果打卡时间在 上课的前40分钟内容 ~ 上课的开始时间后10分内, 认为正常出勤了 返回 0
否则认为迟到出勤 返回 1
情况2: 没有出去, 认为没有来 返回 2
第二类: 指标: 计算 请假人数, 请假率
涉及维度:
时间维度: 年 月 天 上午 下午 晚自习
班级维度:
涉及表:
student_leave_apply: 学生请假表 (事实表)
tbh_class_time_table: 班级作息时间表 (维度表)
course_table_upload_detail: 课程表 (维度表)
表关联条件:
学生请假表.class_id = 班级作息时间表.class_id
学生请假表.class_id = 课程表.class_id
涉及字段:
时间维度: 课程表.class_date
班级维度: 课程表.class_id
指标字段: 请假表.student_id
需要进行去重统计操作
过滤条件:
课程表:
content 不能为空 为null 以及不能为 开班典礼
获取有效的作息时间:
课程表.class_date between 作息表.use_begin_date and 作息表.use_end_date
学生请假表:保证请假数据有效的
audit_state =1 -- 审核通过
cancel_state = 0 -- 没有取消
valid_state = 1 -- 有效的
判断是否请假条件:
请假的开始时间(请假表.begin_time) <= 上课的开始时间 (morning_begin_time |afternon_begin_time | evening_begin_time)
请假的结束时间(请假表.end_time) >= 上课的开始时间(morning_begin_time |afternon_begin_time | evening_begin_time)
第三类: 计算旷课人数, 旷课率
涉及维度:
时间维度: 年 月 天 上午 下午 晚自习
班级维度:
计算标准:
旷课人数 = 当日在读人数 - 出勤人数(正常出勤+迟到出勤) -请假人数业务数据准备工作
create database teach default character set utf8mb4 collate utf8mb4_unicode_ci;
mysql -h 172.16.100.42 -P 4000 -uroot -p'84^j1TZ-+2b9A@S7Hs' teach < teach.sql
-- 表1: 当日在读人数表 class_studying_student_count
-- 表2: 班级课表 course_table_upload_detail
-- 表3: 学生请假表 student_leave_apply
-- 表4: 班级作息时间表 tbh_class_time_table
-- 表5: 学生打卡记录表 tbh_student_signin_record
-- 表6: 日期表 calendar建模分析
* ODS层: 源数据层
作用: 对接数据源, 一般和数据源的表保持相同粒度
一般存放事实表以及少量的维度表
建表方案:
构建两张表 : 学生打卡记录表 和 学生请假表
在构建的时候, 需要额外添加一个分区字段: start_time(抽取数据的时间)
* DIM层: 维度层
作用: 用于存储维度表的数据, 一般和数据源对表保持相同粒度
建表方案:
构建三个表: 课程日期表, 班级作息时间表, 当日在读人数表
在构建的时候, 需要额外添加一个分区字段: start_time(抽取数据的时间)
* DWD层: 明细层
作用: 1) 清洗转换操作 2) 少量的维度退化操作 (不需要)
清洗转换操作:
可做
1) 日期转换为 年 月 日
2) 过滤无效的请假数据
3) 过滤没有开启公屏的数据
目前假设没有这些操作, 所以此时DWD层 是不需要存在的
* DWM层: 中间层
作用: 维度退化 以及 提前聚合操作
处理逻辑: 先 分 在 合的操作
建表方案:
第一个表: 学生出勤状态表
作用: 用于统计每天每个班级, 每个学生的出勤状态(0(正常出勤) 1(迟到出勤) 2(没来))
表字段 :
yearinfo, monthinfo,dayinfo, 班级id, 学生id, 上午出勤, 下午出勤, 晚自习出勤
第二个表: 班级出勤人数表
作用: 用于统计每天每个班级的出勤人数(出勤人数, 迟到出勤人数)
表字段:
yearinfo, monthinfo,dayinfo, 班级id,上午出勤人数, 上午迟到人数, 下午出勤人数, 下午迟到人数, 晚上出勤人数, 晚上迟到人数
第三个表: 班级请假人数表
作用: 用于统计每天每个班级的请假人数
表字段:
yearinfo, monthinfo,dayinfo, 班级id, 上午请假人数, 下午请假人数, 晚上请假人数
第四个表: 班级旷课人数表
作用: 用于统计每天每个班级的旷课人数
表字段:
yearinfo, monthinfo,dayinfo, 班级id, 上午旷课人数, 下午旷课人数, 晚上旷课人数
第五个表: 班级指标汇总表 (提前聚合表)
作用: 用于将前几个表相关的指标数据汇总起来, 同时计算出 相关比率操作 (统计每天的上午 下午 晚自习)
表字段:
yearinfo, monthinfo,dayinfo, 班级id, 班级当日在读人数,
上午出勤人数, 上午出勤率, 下午出勤人数 下午出勤率, 晚上出勤人数, 晚上出勤率
上午迟到人数, 上午迟到率, 下午迟到人数, 下午迟到率, 晚上迟到人数, 晚上迟到率
上午请假人数, 上午请假率, 下午请假人数, 下午请假率, 晚上请假人数, 晚上请假率
上午旷课人数, 上午旷课率, 下午旷课人数, 下午旷课率, 晚上旷课人数, 晚上旷课率
* DWS层: 业务层
作用: 细化维度统计操作
建表方案:
yearinfo, monthinfo,dayinfo, 班级id, 班级当日在读人数, time_str,time_type
上午出勤人数, 上午出勤率, 下午出勤人数 下午出勤率, 晚上出勤人数, 晚上出勤率
上午迟到人数, 上午迟到率, 下午迟到人数, 下午迟到率, 晚上迟到人数, 晚上迟到率
上午请假人数, 上午请假率, 下午请假人数, 下午请假率, 晚上请假人数, 晚上请假率
上午旷课人数, 上午旷课率, 下午旷课人数, 下午旷课率, 晚上旷课人数, 晚上旷课率建模操作
-- ODS层:
-- 学生打卡信息表:
CREATE TABLE IF NOT EXISTS itcast_ods.student_signin_ods (
id int,
normal_class_flag int comment '是否正课 1 正课 2 自习 3 休息',
time_table_id int comment '作息时间id normal_class_flag=2 关联tbh_school_time_table 或者 normal_class_flag=1 关联 tbh_class_time_table',
class_id int comment '班级id',
student_id int comment '学员id',
signin_time String comment '签到时间',
signin_date String comment '签到日期',
inner_flag int comment '内外网标志 0 外网 1 内网',
signin_type int comment '签到类型 1 心跳打卡 2 老师补卡 3 直播打卡',
share_state int comment '共享屏幕状态 0 否 1是 在上午或下午段有共屏记录，则该段所有记录该字段为1，内网默认为1 外网默认为0 (暂不用)',
inner_ip String comment '内网ip地址',
create_time String comment '创建时间')
comment '学生打卡记录表'
PARTITIONED BY (dt STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY','orc.bloom.filter.columns'='time_table_id,class_id,signin_date,share_state');
-- 学生请假信息表:
CREATE TABLE IF NOT EXISTS itcast_ods.student_leave_apply_ods (
id int,
class_id int comment '班级id',
student_id int comment '学员id',
audit_state int comment '审核状态 0 待审核 1 通过 2 不通过',
audit_person int comment '审核人',
audit_time String comment '审核时间',
audit_remark String comment '审核备注',
leave_type int comment '请假类型 1 请假 2 销假 （查询是否请假不用过滤此类型，通过有效状态来判断）',
leave_reason int comment '请假原因 1 事假 2 病假',
begin_time String comment '请假开始时间',
begin_time_type int comment '1：上午 2：下午 3：晚自习',
end_time String comment '请假结束时间',
end_time_type int comment '1：上午 2：下午 3：晚自习',
days float comment '请假/已休天数',
cancel_state int comment '撤销状态 0 未撤销 1 已撤销',
cancel_time String comment '撤销时间',
old_leave_id int comment '原请假id，只有leave_type =2 销假的时候才有',
leave_remark String comment '请假/销假说明',
valid_state int comment '是否有效（0：无效 1：有效）',
create_time String comment '创建时间')
comment '学生请假申请表'
PARTITIONED BY (dt STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY','orc.bloom.filter.columns'='class_id,audit_state,cancel_state,valid_state');
-- DIM层:
-- 日期课程表
CREATE TABLE IF NOT EXISTS itcast_dimen.course_table_upload_detail_dimen
(
id int comment 'id',
base_id int comment '课程主表id',
class_id int comment '班级id',
class_date STRING comment '上课日期',
content STRING comment '课程内容',
teacher_id int comment '老师id',
teacher_name STRING comment '老师名字',
job_number STRING comment '工号',
classroom_id int comment '教室id',
classroom_name STRING comment '教室名称',
is_outline int comment '是否大纲 0 否 1 是',
class_mode int comment '上课模式 0 传统全天 1 AB上午 2 AB下午 3 线上直播',
is_stage_exam int comment '是否阶段考试（0：否 1：是）',
is_pay int comment '代课费（0：无 1：有）',
tutor_teacher_id int comment '晚自习辅导老师id',
tutor_teacher_name STRING comment '辅导老师姓名',
tutor_job_number STRING comment '晚自习辅导老师工号',
is_subsidy int comment '晚自习补贴（0：无 1：有）',
answer_teacher_id int comment '答疑老师id',
answer_teacher_name STRING comment '答疑老师姓名',
answer_job_number STRING comment '答疑老师工号',
remark STRING comment '备注',
create_time STRING comment '创建时间')
comment '班级课表明细表'
PARTITIONED BY (dt STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY','orc.bloom.filter.columns'='class_id,class_date');
-- 班级作息时间表
CREATE TABLE IF NOT EXISTS itcast_dimen.class_time_dimen (
id int,
class_id int comment '班级id',
morning_template_id int comment '上午出勤模板id',
morning_begin_time STRING comment '上午开始时间',
morning_end_time STRING comment '上午结束时间',
afternoon_template_id int comment '下午出勤模板id',
afternoon_begin_time STRING comment '下午开始时间',
afternoon_end_time STRING comment '下午结束时间',
evening_template_id int comment '晚上出勤模板id',
evening_begin_time STRING comment '晚上开始时间',
evening_end_time STRING comment '晚上结束时间',
use_begin_date STRING comment '使用开始日期',
use_end_date STRING comment '使用结束日期',
create_time STRING comment '创建时间',
create_person int comment '创建人',
remark STRING comment '备注')
comment '班级作息时间表'
PARTITIONED BY (dt STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY','orc.bloom.filter.columns'='id,class_id');
-- 当日在读人数表:
CREATE TABLE IF NOT EXISTS itcast_dimen.class_studying_student_count_dimen (
id int,
school_id int comment '校区id',
subject_id int comment '学科id',
class_id int comment '班级id',
studying_student_count int comment '在读班级人数',
studying_date STRING comment '在读日期')
comment '在读班级的每天在读学员人数'
PARTITIONED BY (dt STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY','orc.bloom.filter.columns'='studying_student_count,studying_date');
-- DWM层:
-- 学生出勤状态信息表
CREATE TABLE IF NOT EXISTS itcast_dwm.student_attendance_dwm (
dateinfo String comment '日期',
class_id int comment '班级id',
student_id int comment '学员id',
morning_att String comment '上午出勤情况：0.正常出勤、1.迟到、2.其他（请假+旷课）',
afternoon_att String comment '下午出勤情况：0.正常出勤、1.迟到、2.其他（请假+旷课）',
evening_att String comment '晚自习出勤情况：0.正常出勤、1.迟到、2.其他（请假+旷课）')
comment '学生出勤(正常出勤和迟到)数据'
PARTITIONED BY (yearinfo STRING, monthinfo STRING, dayinfo STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY');
-- 班级出勤人数表:
CREATE TABLE IF NOT EXISTS itcast_dwm.class_attendance_dwm (
dateinfo String comment '日期',
class_id int comment '班级id',
morning_att_count String comment '上午出勤人数',
afternoon_att_count String comment '下午出勤人数',
evening_att_count String comment '晚自习出勤人数',
morning_late_count String comment '上午迟到人数',
afternoon_late_count String comment '下午迟到人数',
evening_late_count String comment '晚自习迟到人数')
comment '学生出勤(正常出勤和迟到)数据'
PARTITIONED BY (yearinfo STRING, monthinfo STRING, dayinfo STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY');
-- 班级请假人数表
CREATE TABLE IF NOT EXISTS itcast_dwm.class_leave_dwm (
dateinfo String comment '日期',
class_id int comment '班级id',
morning_leave_count String comment '上午请假人数',
afternoon_leave_count String comment '下午请假人数',
evening_leave_count String comment '晚自习请假人数')
comment '班级请假数据统计'
PARTITIONED BY (yearinfo STRING, monthinfo STRING, dayinfo STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY');
-- 班级旷课人数表
CREATE TABLE IF NOT EXISTS itcast_dwm.class_truant_dwm (
dateinfo String comment '日期',
class_id int comment '班级id',
morning_truant_count String comment '上午旷课人数',
afternoon_truant_count String comment '下午旷课人数',
evening_truant_count String comment '晚自习旷课人数')
comment '班级请假数据统计'
PARTITIONED BY (yearinfo STRING, monthinfo STRING, dayinfo STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY');
-- 汇总表
CREATE TABLE IF NOT EXISTS itcast_dwm.class_all_dwm (
dateinfo String COMMENT '日期',
class_id INT COMMENT '班级id',
studying_student_count INT COMMENT '在读班级人数',
morning_att_count String COMMENT '上午出勤人数',
morning_att_ratio String COMMENT '上午出勤率',
afternoon_att_count String COMMENT '下午出勤人数',
afternoon_att_ratio String COMMENT '下午出勤率',
evening_att_count String COMMENT '晚自习出勤人数',
evening_att_ratio String COMMENT '晚自习出勤率',
morning_late_count String COMMENT '上午迟到人数',
morning_late_ratio String COMMENT '上午迟到率',
afternoon_late_count String COMMENT '下午迟到人数',
afternoon_late_ratio String COMMENT '下午迟到率',
evening_late_count String COMMENT '晚自习迟到人数',
evening_late_ratio String COMMENT '晚自习迟到率',
morning_leave_count String COMMENT '上午请假人数',
morning_leave_ratio String COMMENT '上午请假率',
afternoon_leave_count String COMMENT '下午请假人数',
afternoon_leave_ratio String COMMENT '下午请假率',
evening_leave_count String COMMENT '晚自习请假人数',
evening_leave_ratio String COMMENT '晚自习请假率',
morning_truant_count String COMMENT '上午旷课人数',
morning_truant_ratio String COMMENT '上午旷课率',
afternoon_truant_count String COMMENT '下午旷课人数',
afternoon_truant_ratio String COMMENT '下午旷课率',
evening_truant_count String COMMENT '晚自习旷课人数',
evening_truant_ratio String COMMENT '晚自习旷课率')
comment '班级请假数据统计'
PARTITIONED BY (yearinfo STRING, monthinfo STRING, dayinfo STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
stored as orcfile
TBLPROPERTIES ('orc.compress'='SNAPPY');
-- DWS层:
CREATE TABLE IF NOT EXISTS itcast_dws.class_attendance_dws (
dateinfo String COMMENT '日期',
class_id INT COMMENT '班级id',
studying_student_count INT COMMENT '在读班级人数',
morning_att_count String COMMENT '上午出勤人数',
morning_att_ratio String COMMENT '上午出勤率',
afternoon_att_count String COMMENT '下午出勤人数',
afternoon_att_ratio String COMMENT '下午出勤率',
evening_att_count String COMMENT '晚自习出勤人数',
evening_att_ratio String COMMENT '晚自习出勤率',
morning_late_count String COMMENT '上午迟到人数',
morning_late_ratio String COMMENT '上午迟到率',
afternoon_late_count String COMMENT '下午迟到人数',
afternoon_late_ratio String COMMENT '下午迟到率',
evening_late_count String COMMENT '晚自习迟到人数',
evening_late_ratio String COMMENT '晚自习迟到率',
morning_leave_count String COMMENT '上午请假人数',
morning_leave_ratio String COMMENT '上午请假率',
afternoon_leave_count String COMMENT '下午请假人数',
afternoon_leave_ratio String COMMENT '下午请假率',
evening_leave_count String COMMENT '晚自习请假人数',
evening_leave_ratio String COMMENT '晚自习请假率',
morning_truant_count String COMMENT '上午旷课人数',
morning_truant_ratio String COMMENT '上午旷课率',
afternoon_truant_count String COMMENT '下午旷课人数',
afternoon_truant_ratio String COMMENT '下午旷课率',
evening_truant_count String COMMENT '晚自习旷课人数',
evening_truant_ratio String COMMENT '晚自习旷课率',
time_type STRING COMMENT '聚合时间类型：1、按小时聚合；2、按天聚合；3、按周聚合；4、按月聚合；5、按年聚合。'
)
COMMENT '班级请假数据统计'
PARTITIONED BY ( yearinfo STRING, monthinfo STRING, dayinfo STRING )
ROW FORMAT DELIMITED FIELDS TERMINATED BY '\t'
stored AS orcfile
TBLPROPERTIES ( 'orc.compress' = 'SNAPPY' );数据采集操作
ODS层数据采集操作
# 学生打卡记录表
sqoop import \
--connect jdbc:mysql://172.16.100.42:4000/teach \
--username root \
--password 84^j1TZ-+2b9A@S7Hs \
--query 'select *,"2021-10-07" as dt from tbh_student_signin_record where 1=1 and $CONDITIONS' \
--hcatalog-database itcast_ods \
--hcatalog-table student_signin_ods \
-m 1
# 学生请假信息表
sqoop import \
--connect jdbc:mysql://172.16.100.42:4000/teach \
--username root \
--password 84^j1TZ-+2b9A@S7Hs \
--query 'select *,"2021-10-07" as dt from student_leave_apply where 1=1 and $CONDITIONS' \
--hcatalog-database itcast_ods \
--hcatalog-table student_leave_apply_ods \
-m 1DIM层:
# 当日在读人数表
sqoop import \
--connect jdbc:mysql://172.16.100.42:4000/teach \
--username root \
--password 84^j1TZ-+2b9A@S7Hs \
--query 'select *,"2021-10-07" as dt from class_studying_student_count where 1=1 and $CONDITIONS' \
--hcatalog-database itcast_dimen \
--hcatalog-table class_studying_student_count_dimen \
-m 1
# 日期课程表
sqoop import \
--connect jdbc:mysql://172.16.100.42:4000/teach \
--username root \
--password 84^j1TZ-+2b9A@S7Hs \
--query 'select *,"2021-10-07" as dt from course_table_upload_detail where 1=1 and $CONDITIONS' \
--hcatalog-database itcast_dimen \
--hcatalog-table course_table_upload_detail_dimen \
-m 1
# 作息时间表
sqoop import \
--connect jdbc:mysql://172.16.100.42:4000/teach \
--username root \
--password 84^j1TZ-+2b9A@S7Hs \
--query 'select *,"2021-10-07" as dt from tbh_class_time_table where 1=1 and $CONDITIONS' \
--hcatalog-database itcast_dimen \
--hcatalog-table class_time_dimen \
-m 1数据清洗转换操作
目的: 主要是用于从ODS以及DIM层 将数据灌入到DWM层操作
* 生成学生出勤状态信息表
涉及表:
course_table_upload_detail: 日志课程明细表 (课表) (维度表)
tbh_student_signin_record: 学生打卡记录表 (事实表)
tbh_class_time_table: 班级作息时间表 (维度表)
关联条件:
学生打卡表.class_id = 课程表.class_id
班级作息时间表.id = 学生打卡表.time_table_id
涉及到字段:
时间维度: 课程表.class_date
班级维度: 课程表.class_id
学生维度: 学生打卡表.student_id
指标字段: 学生打卡表.signin_time(打卡时间)
计算方案:
先判断是否出勤
情况1: 出勤了, 再次判断是否是正常出勤和迟到出勤
情况2: 未出勤, 认为没有来
指标判断指标:
作息时间表:
morning_begin_time
morning_end_time
afternoon_begin_time
afternoon_end_time
evening_begin_time
evening_end_time
过滤操作
1) ifnull(ctud.content,'') != '' 相当于 ctud.content != null and ctud.content != ''
转换为hive的操作
nvl(ctud.content,'') != ''
2) 将content中为开班典礼数据过滤掉
ctud.content != '开班典礼'
3) 确保打卡表中学生数据都是开启公屏
学生打卡表.share_state=1
* 基本SQL的实现: 相关的出勤状态未完成
select
ctudd.class_date,
ctudd.class_id,
student_id,
'' as morning_att, --未实现
'' as afternoon_att, --未实现
'' as evening_att, --未实现
substr(ctudd.class_date,1,4) as yearinfo,
substr(ctudd.class_date,6,2) as monthinfo,
substr(ctudd.class_date,9,2) as dayinfo
from (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content !='开班典礼') ctudd
left join (select * from itcast_ods.student_signin_ods where share_state = 1) sso on sso.class_id = ctudd.class_id
left join itcast_dimen.class_time_dimen ctd on ctd.id = sso.time_table_id
group by ctudd.class_date , ctudd.class_id,sso.student_id;
探讨如何判断学生出勤状态呢?
基于学生的打卡时间 以上午为例
如果学生的打卡时间 在 上午上课开始时间前40分钟内 ~~ 上午上课截止时间内
认为 学生出勤了
此时接着判断, 如何学生的打卡时间在 上午上课开始时间前40分钟内容 ~~ 上午上课开始时间后10分钟内
认为 学生是正常出勤 返回 0
否则 认为学生是迟到出勤 返回 1
否则认为学生没有出勤, 直接返回 2
* 探讨. 如何实现日期的 相加 和 相减 (对分钟处理)
在hive中, 并没有发现可以对分钟加减的函数, 只有对天的加减函数, 但是不符合要求, 如何解决呢?
可以尝试将日期数据转换为时间戳, 然后对时间戳进行加减处理 即可解决问题
select unix_timestamp('2021-10-08 15:40:30','yyyy-MM-dd HH:mm:ss') - 40*60;-- 将SQL的结果灌入到DWM层
-- 自动将适合使用MapJoin的JOIN操作转换为MapJoin,默认为true
set hive.auto.convert.join=false;
--分区
SET hive.exec.dynamic.partition=true;
SET hive.exec.dynamic.partition.mode=nonstrict;
set hive.exec.max.dynamic.partitions.pernode=10000;
set hive.exec.max.dynamic.partitions=100000;
set hive.exec.max.created.files=150000;
--hive压缩
set hive.exec.compress.intermediate=true;
set hive.exec.compress.output=true;
--写入时压缩生效
set hive.exec.orc.compression.strategy=COMPRESSION;
--分桶
--set hive.enforce.bucketing=true;
--set hive.enforce.sorting=true;
--set hive.optimize.bucketmapjoin = true;
--set hive.auto.convert.sortmerge.join=true;
--set hive.auto.convert.sortmerge.join.noconditionaltask=true;
--并行执行
set hive.exec.parallel=true;
set hive.exec.parallel.thread.number=8;
--小文件合并
-- set mapred.max.split.size=2147483648;
-- set mapred.min.split.size.per.node=1000000000;
-- set mapred.min.split.size.per.rack=1000000000;
--矢量化查询
set hive.vectorized.execution.enabled=true;
--关联优化器
set hive.optimize.correlation=true;
--读取零拷贝
set hive.exec.orc.zerocopy=true;
--join数据倾斜
set hive.optimize.skewjoin=true;
-- set hive.skewjoin.key=100000;
set hive.optimize.skewjoin.compiletime=true;
set hive.optimize.union.remove=true;
-- group倾斜
set hive.groupby.skewindata=true;
-- 实现上午的学生出勤统计SQL:
insert into table itcast_dwm.student_attendance_dwm partition(yearinfo,monthinfo,dayinfo)
select
ctudd.class_date,
ctudd.class_id,
student_id,
if(
sum(
if(
unix_timestamp(sso.signin_time,'yyyy-MM-dd HH:mm:ss')
between unix_timestamp(concat(ctudd.class_date,' ',ctd.morning_begin_time),'yyyy-MM-dd HH:mm:ss') - 40*60
and unix_timestamp(concat(ctudd.class_date,' ',ctd.morning_end_time),'yyyy-MM-dd HH:mm:ss')
,1,0)
) > 0, -- 如果大于0 认为当天的打卡记录中, 一定是有出勤的打卡记录, 如果小于等于0 认为没有出勤
if(
sum(
if(
unix_timestamp(sso.signin_time,'yyyy-MM-dd HH:mm:ss')
between unix_timestamp(concat(ctudd.class_date,' ',ctd.morning_begin_time),'yyyy-MM-dd HH:mm:ss') - 40*60
and unix_timestamp(concat(ctudd.class_date,' ',ctd.morning_begin_time),'yyyy-MM-dd HH:mm:ss') + 10*60
,1,0)
) > 0, -- 如果大于0, 认为当天打卡记录中, 一定是有正常出勤的记录, 否则认为迟到出勤
0,1),
2)as morning_att,
if(
sum(
if(
unix_timestamp(sso.signin_time,'yyyy-MM-dd HH:mm:ss')
between unix_timestamp(concat(ctudd.class_date,' ',ctd.afternoon_begin_time),'yyyy-MM-dd HH:mm:ss') - 40*60
and unix_timestamp(concat(ctudd.class_date,' ',ctd.afternoon_end_time),'yyyy-MM-dd HH:mm:ss')
,1,0)
) > 0, -- 如果大于0 认为当天的打卡记录中, 一定是有出勤的打卡记录, 如果小于等于0 认为没有出勤
if(
sum(
if(
unix_timestamp(sso.signin_time,'yyyy-MM-dd HH:mm:ss')
between unix_timestamp(concat(ctudd.class_date,' ',ctd.afternoon_begin_time),'yyyy-MM-dd HH:mm:ss') - 40*60
and unix_timestamp(concat(ctudd.class_date,' ',ctd.afternoon_begin_time),'yyyy-MM-dd HH:mm:ss') + 10*60
, 1 , 0)
) >0, -- 如果大于0, 认为当天打卡记录中, 一定是有正常出勤的记录, 否则认为迟到出勤
0,1),
2) as afternoon_att,
if(
sum(
if(
unix_timestamp(sso.signin_time,'yyyy-MM-dd HH:mm:ss')
between unix_timestamp(concat(ctudd.class_date,' ',ctd.evening_begin_time),'yyyy-MM-dd HH:mm:ss') - 40*60
and unix_timestamp(concat(ctudd.class_date,' ',ctd.evening_end_time),'yyyy-MM-dd HH:mm:ss')
,1,0)
) > 0, -- 如果大于0 认为当天的打卡记录中, 一定是有出勤的打卡记录, 如果小于等于0 认为没有出勤
if(
sum(
if(
unix_timestamp(sso.signin_time,'yyyy-MM-dd HH:mm:ss')
between unix_timestamp(concat(ctudd.class_date,' ',ctd.evening_begin_time),'yyyy-MM-dd HH:mm:ss') - 40*60
and unix_timestamp(concat(ctudd.class_date,' ',ctd.evening_begin_time),'yyyy-MM-dd HH:mm:ss') + 10*60
, 1 , 0)
) >0, -- 如果大于0, 认为当天打卡记录中, 一定是有正常出勤的记录, 否则认为迟到出勤
0,1),
2) as evening_att, --未实现
substr(ctudd.class_date,1,4) as yearinfo,
substr(ctudd.class_date,6,2) as monthinfo,
substr(ctudd.class_date,9,2) as dayinfo
from (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content !='开班典礼') ctudd
left join (select * from itcast_ods.student_signin_ods where share_state = 1) sso on sso.class_id = ctudd.class_id
left join itcast_dimen.class_time_dimen ctd on ctd.id = sso.time_table_id
where ctudd.class_date in ('2019-09-03','2019-09-04','2019-09-05')
group by ctudd.class_date , ctudd.class_id,sso.student_id;
-- 班级出勤人数表
insert into table itcast_dwm.class_attendance_dwm partition(yearinfo,monthinfo,dayinfo)
select
dateinfo,
class_id,
count(
case
when morning_att in ('0','1') then student_id
else null end
) as morning_att_count,
count(
case
when afternoon_att in ('0','1') then student_id
else null end
) as afternoon_att_count,
count(
case
when evening_att in ('0','1') then student_id
else null end
) as evening_att_count,
sum(
case
when morning_att ='1' then 1
else 0 end
) as morning_late_count,
sum(
case
when afternoon_att ='1' then 1
else 0 end
) as afternoon_late_count,
sum(
case
when evening_att ='1' then 1
else 0 end
) as evening_late_count,
yearinfo,
monthinfo,
dayinfo
from itcast_dwm.student_attendance_dwm
group by dateinfo,yearinfo,monthinfo,dayinfo,class_id;班级请假人数表的统计生成
涉及表:
student_leave_apply: 学生请假表 (事实表)
tbh_class_time_table: 班级作息时间表 (维度表)
course_table_upload_detail: 课程表 (维度表)
表关联条件:
学生请假表.class_id = 班级作息时间表.class_id
学生请假表.class_id = 课程表.class_id
涉及字段:
时间维度: 课程表.class_date
班级维度: 课程表.class_id
指标字段: 请假表.student_id
需要进行去重统计操作
过滤条件:
课程表:
content 不能为空 为null 以及不能为 开班典礼
获取有效的作息时间:
课程表.class_date between 作息表.use_begin_date and 作息表.use_end_date
学生请假表:保证请假数据有效的
audit_state =1 -- 审核通过
cancel_state = 0 -- 没有取消
valid_state = 1 -- 有效的
判断是否请假条件:
请假的开始时间(请假表.begin_time) <= 上课的开始时间 (morning_begin_time |afternon_begin_time | evening_begin_time)
请假的结束时间(请假表.end_time) >= 上课的开始时间(morning_begin_time |afternon_begin_time | evening_begin_time)-- 班级请假人数表的统计生成
-- 模拟一条请假数据
INSERT INTO itcast_ods.student_leave_apply_ods partition (dt)
values (125, 5032, 119142, 1, 3491, '2019-09-08 16:42:29', '狂犬疫苗最后一针', 1, 2, '2019-09-03 08:00:00', 2, '2019-09-03 23:30:00', 2, 1, 0, null, null, '医院打针，狂犬疫苗最后一针了', 1, '2019-09-02 08:56:54','2020-07-07');
-- 计算上午的每天各个班级请假人数
select
ctudd.class_date,
ctudd.class_id,
count(distinct sla.student_id) as morning_leave_count
from (select * from itcast_ods.student_leave_apply_ods where audit_state = 1 and cancel_state = 0 and valid_state = 1) sla
join itcast_dimen.class_time_dimen ctd on sla.class_id = ctd.class_id
join (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd on ctudd.class_id = sla.class_id
where ctudd.class_date between ctd.use_begin_date and ctd.use_end_date
and concat(ctudd.class_date,' ',ctd.morning_begin_time) >= sla.begin_time
and concat(ctudd.class_date,' ',ctd.morning_begin_time) <= sla.end_time
group by ctudd.class_date,ctudd.class_id;
-- 计算下午的每天各个班级的请假人数
select
ctudd.class_date,
ctudd.class_id,
count(distinct sla.student_id) as afternoon_leave_count
from (select * from itcast_ods.student_leave_apply_ods where audit_state = 1 and cancel_state = 0 and valid_state = 1) sla
join itcast_dimen.class_time_dimen ctd on sla.class_id = ctd.class_id
join (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd on ctudd.class_id = sla.class_id
where ctudd.class_date between ctd.use_begin_date and ctd.use_end_date
and concat(ctudd.class_date,' ',ctd.afternoon_begin_time) >= sla.begin_time
and concat(ctudd.class_date,' ',ctd.afternoon_begin_time) <= sla.end_time
group by ctudd.class_date,ctudd.class_id;
-- 计算晚自习的每天各个班级的请假人数
select
ctudd.class_date,
ctudd.class_id,
count(distinct sla.student_id) as evening_leave_count
from (select * from itcast_ods.student_leave_apply_ods where audit_state = 1 and cancel_state = 0 and valid_state = 1) sla
join itcast_dimen.class_time_dimen ctd on sla.class_id = ctd.class_id
join (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd on ctudd.class_id = sla.class_id
where ctudd.class_date between ctd.use_begin_date and ctd.use_end_date
and concat(ctudd.class_date,' ',ctd.evening_begin_time) >= sla.begin_time
and concat(ctudd.class_date,' ',ctd.evening_begin_time) <= sla.end_time
group by ctudd.class_date,ctudd.class_id;
-- 合并数据 全外连接(满外) full join
-- 将最终合并的结果灌入到DWM层
-- 自动将适合使用MapJoin的JOIN操作转换为MapJoin,默认为true
set hive.auto.convert.join=false;
--分区
SET hive.exec.dynamic.partition=true;
SET hive.exec.dynamic.partition.mode=nonstrict;
set hive.exec.max.dynamic.partitions.pernode=10000;
set hive.exec.max.dynamic.partitions=100000;
set hive.exec.max.created.files=150000;
--hive压缩
set hive.exec.compress.intermediate=true;
set hive.exec.compress.output=true;
--写入时压缩生效
set hive.exec.orc.compression.strategy=COMPRESSION;
--分桶
--set hive.enforce.bucketing=true;
--set hive.enforce.sorting=true;
--set hive.optimize.bucketmapjoin = true;
--set hive.auto.convert.sortmerge.join=true;
--set hive.auto.convert.sortmerge.join.noconditionaltask=true;
--并行执行
set hive.exec.parallel=true;
set hive.exec.parallel.thread.number=8;
--小文件合并
-- set mapred.max.split.size=2147483648;
-- set mapred.min.split.size.per.node=1000000000;
-- set mapred.min.split.size.per.rack=1000000000;
--矢量化查询
set hive.vectorized.execution.enabled=true;
--关联优化器
set hive.optimize.correlation=true;
--读取零拷贝
set hive.exec.orc.zerocopy=true;
--join数据倾斜
set hive.optimize.skewjoin=false;
-- set hive.skewjoin.key=100000;
set hive.optimize.skewjoin.compiletime=false;
set hive.optimize.union.remove=false;
-- group倾斜
set hive.groupby.skewindata=false;
with A as (select
ctudd.class_date,
ctudd.class_id,
count(distinct sla.student_id) as morning_leave_count
from (select * from itcast_ods.student_leave_apply_ods where audit_state = 1 and cancel_state = 0 and valid_state = 1) sla
join itcast_dimen.class_time_dimen ctd on sla.class_id = ctd.class_id
join (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd on ctudd.class_id = sla.class_id
where ctudd.class_date between ctd.use_begin_date and ctd.use_end_date
and concat(ctudd.class_date,' ',ctd.morning_begin_time) >= sla.begin_time
and concat(ctudd.class_date,' ',ctd.morning_begin_time) <= sla.end_time
group by ctudd.class_date,ctudd.class_id),
B AS (select
ctudd.class_date,
ctudd.class_id,
count(distinct sla.student_id) as afternoon_leave_count
from (select * from itcast_ods.student_leave_apply_ods where audit_state = 1 and cancel_state = 0 and valid_state = 1) sla
join itcast_dimen.class_time_dimen ctd on sla.class_id = ctd.class_id
join (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd on ctudd.class_id = sla.class_id
where ctudd.class_date between ctd.use_begin_date and ctd.use_end_date
and concat(ctudd.class_date,' ',ctd.afternoon_begin_time) >= sla.begin_time
and concat(ctudd.class_date,' ',ctd.afternoon_begin_time) <= sla.end_time
group by ctudd.class_date,ctudd.class_id),
C AS (select
ctudd.class_date,
ctudd.class_id,
count(distinct sla.student_id) as evening_leave_count
from (select * from itcast_ods.student_leave_apply_ods where audit_state = 1 and cancel_state = 0 and valid_state = 1) sla
join itcast_dimen.class_time_dimen ctd on sla.class_id = ctd.class_id
join (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd on ctudd.class_id = sla.class_id
where ctudd.class_date between ctd.use_begin_date and ctd.use_end_date
and concat(ctudd.class_date,' ',ctd.evening_begin_time) >= sla.begin_time
and concat(ctudd.class_date,' ',ctd.evening_begin_time) <= sla.end_time
group by ctudd.class_date,ctudd.class_id),
temp as (select
coalesce(A.class_date,B.class_date,C.class_date) AS class_date,
coalesce(A.class_id,B.class_id,C.class_id) AS class_id,
nvl(A.morning_leave_count,0) as morning_leave_count,
nvl(B.afternoon_leave_count,0) as afternoon_leave_count,
nvl(C.evening_leave_count,0) as evening_leave_count
from A
full join B on A.class_date = B.class_date and A.class_id = B.class_id
full join C on A.class_date = C.class_date and A.class_id = C.class_id)
insert into table itcast_dwm.class_leave_dwm partition(yearinfo,monthinfo,dayinfo)
select
class_date,
class_id,
sum(morning_leave_count) as morning_leave_count,
sum(afternoon_leave_count) as afternoon_leave_count,
sum(evening_leave_count) as evening_leave_count,
substr(class_date,1,4) as yearinfo,
substr(class_date,6,2) as monthinfo,
substr(class_date,9,2) as dayinfo
from temp group by class_date,class_id;旷课人数表统计
计算规则:
- 旷课人数 = 当日在读人数 - 出勤人数 - 请假人数
-- 自动将适合使用MapJoin的JOIN操作转换为MapJoin,默认为true
set hive.auto.convert.join=false;
--分区
SET hive.exec.dynamic.partition=true;
SET hive.exec.dynamic.partition.mode=nonstrict;
set hive.exec.max.dynamic.partitions.pernode=10000;
set hive.exec.max.dynamic.partitions=100000;
set hive.exec.max.created.files=150000;
--hive压缩
set hive.exec.compress.intermediate=true;
set hive.exec.compress.output=true;
--写入时压缩生效
set hive.exec.orc.compression.strategy=COMPRESSION;
--分桶
--set hive.enforce.bucketing=true;
--set hive.enforce.sorting=true;
--set hive.optimize.bucketmapjoin = true;
--set hive.auto.convert.sortmerge.join=true;
--set hive.auto.convert.sortmerge.join.noconditionaltask=true;
--并行执行
set hive.exec.parallel=true;
set hive.exec.parallel.thread.number=8;
--小文件合并
-- set mapred.max.split.size=2147483648;
-- set mapred.min.split.size.per.node=1000000000;
-- set mapred.min.split.size.per.rack=1000000000;
--矢量化查询
set hive.vectorized.execution.enabled=true;
--关联优化器
set hive.optimize.correlation=true;
--读取零拷贝
set hive.exec.orc.zerocopy=true;
--join数据倾斜
set hive.optimize.skewjoin=false;
-- set hive.skewjoin.key=100000;
set hive.optimize.skewjoin.compiletime=false;
set hive.optimize.union.remove=false;
-- group倾斜
set hive.groupby.skewindata=false;
insert into table itcast_dwm.class_truant_dwm partition(yearinfo,monthinfo,dayinfo)
select
ctudd.class_date as dateinfo,
ctudd.class_id,
cssc.studying_student_count - nvl(cad.morning_att_count,0) - nvl(cld.morning_leave_count,0) as morning_truant_count,
cssc.studying_student_count - nvl(cad.afternoon_att_count,0) - nvl(cld.afternoon_leave_count,0) as afternoon_truant_count,
cssc.studying_student_count - nvl(cad.evening_att_count,0) - nvl(cld.evening_leave_count,0) as evening_truant_count,
substr(ctudd.class_date,1,4) as yearinfo,
substr(ctudd.class_date,6,2) as monthinfo,
substr(ctudd.class_date,9,2) as dayinfo
from (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd
left join itcast_dimen.class_studying_student_count_dimen cssc on ctudd.class_date = cssc.studying_date and ctudd.class_id = cssc.class_id
left join itcast_dwm.class_attendance_dwm cad on ctudd.class_id = cad.class_id and ctudd.class_date = cad.dateinfo
left join itcast_dwm.class_leave_dwm cld on ctudd.class_id = cld.class_id and ctudd.class_date = cld.dateinfo
where ctudd.class_date in('2019-09-03','2019-09-04','2019-09-05')考勤汇总统计
-- 自动将适合使用MapJoin的JOIN操作转换为MapJoin,默认为true
set hive.auto.convert.join=false;
--分区
SET hive.exec.dynamic.partition=true;
SET hive.exec.dynamic.partition.mode=nonstrict;
set hive.exec.max.dynamic.partitions.pernode=10000;
set hive.exec.max.dynamic.partitions=100000;
set hive.exec.max.created.files=150000;
--hive压缩
set hive.exec.compress.intermediate=true;
set hive.exec.compress.output=true;
--写入时压缩生效
set hive.exec.orc.compression.strategy=COMPRESSION;
--分桶
--set hive.enforce.bucketing=true;
--set hive.enforce.sorting=true;
--set hive.optimize.bucketmapjoin = true;
--set hive.auto.convert.sortmerge.join=true;
--set hive.auto.convert.sortmerge.join.noconditionaltask=true;
--并行执行
set hive.exec.parallel=true;
set hive.exec.parallel.thread.number=8;
--小文件合并
-- set mapred.max.split.size=2147483648;
-- set mapred.min.split.size.per.node=1000000000;
-- set mapred.min.split.size.per.rack=1000000000;
--矢量化查询
set hive.vectorized.execution.enabled=true;
--关联优化器
set hive.optimize.correlation=true;
--读取零拷贝
set hive.exec.orc.zerocopy=true;
--join数据倾斜
set hive.optimize.skewjoin=false;
-- set hive.skewjoin.key=100000;
set hive.optimize.skewjoin.compiletime=false;
set hive.optimize.union.remove=false;
-- group倾斜
set hive.groupby.skewindata=false;
insert OVERWRITE table itcast_dwm.class_all_dwm partition(yearinfo,monthinfo,dayinfo)
select
ctudd.class_date as dateinfo,
ctudd.class_id,
cssc.studying_student_count,
cad.morning_att_count,
concat(round(nvl(cad.morning_att_count,0) / cssc.studying_student_count * 100,2),'%') as morning_att_ratio,
cad.afternoon_att_count,
concat(round(nvl(cad.afternoon_att_count,0) / cssc.studying_student_count * 100,2),'%') as afternoon_att_ratio,
cad.evening_att_count,
concat(round(nvl(cad.evening_att_count,0) / cssc.studying_student_count * 100,2),'%') as evening_att_ratio,
cad.morning_late_count,
concat(round(nvl(cad.morning_late_count,0) / cssc.studying_student_count * 100,2),'%') as morning_late_ratio,
cad.afternoon_late_count,
concat(round(nvl(cad.afternoon_late_count,0) / cssc.studying_student_count * 100,2),'%') as afternoon_late_ratio,
cad.evening_late_count,
concat(round(nvl(cad.evening_late_count,0) / cssc.studying_student_count * 100,2),'%') as evening_late_ratio,
cld.morning_leave_count,
concat(round(nvl(cld.morning_leave_count,0) / cssc.studying_student_count * 100,2),'%') as morning_leave_ratio,
cld.afternoon_leave_count,
concat(round(nvl(cld.afternoon_leave_count,0) / cssc.studying_student_count * 100,2),'%') as afternoon_leave_ratio,
cld.evening_leave_count,
concat(round(nvl(cld.evening_leave_count,0) / cssc.studying_student_count * 100,2),'%') as evening_leave_ratio,
ctd.morning_truant_count,
concat(round(nvl(ctd.morning_truant_count,0) / cssc.studying_student_count * 100,2),'%') as morning_truant_ratio,
ctd.afternoon_truant_count,
concat(round(nvl(ctd.afternoon_truant_count,0) / cssc.studying_student_count * 100,2),'%') as afternoon_truant_ratio,
ctd.evening_truant_count,
concat(round(nvl(ctd.evening_truant_count,0) / cssc.studying_student_count * 100,2),'%') as evening_truant_ratio,
substr(ctudd.class_date,1,4) as yearinfo,
substr(ctudd.class_date,6,2) as monthinfo,
substr(ctudd.class_date,9,2) as dayinfo
from (select * from itcast_dimen.course_table_upload_detail_dimen where nvl(content,'')!='' and content != '开班典礼') ctudd
left join itcast_dimen.class_studying_student_count_dimen cssc on cssc.class_id = ctudd.class_id and cssc.studying_date = ctudd.class_date
left join itcast_dwm.class_attendance_dwm cad on ctudd.class_id = cad.class_id and ctudd.class_date = cad.dateinfo
left join itcast_dwm.class_leave_dwm cld on ctudd.class_id = cld.class_id and ctudd.class_date = cld.dateinfo
left join itcast_dwm.class_truant_dwm ctd on ctudd.class_id = ctd.class_id and ctudd.class_date = ctd.dateinfo
where ctudd.class_date in ('2019-09-03','2019-09-04','2019-09-05');数据的统计分析(DWS)
DWS: 细化维度统计操作
-- 自动将适合使用MapJoin的JOIN操作转换为MapJoin,默认为true
set hive.auto.convert.join=false;
--分区
SET hive.exec.dynamic.partition=true;
SET hive.exec.dynamic.partition.mode=nonstrict;
set hive.exec.max.dynamic.partitions.pernode=10000;
set hive.exec.max.dynamic.partitions=100000;
set hive.exec.max.created.files=150000;
--hive压缩
set hive.exec.compress.intermediate=true;
set hive.exec.compress.output=true;
--写入时压缩生效
set hive.exec.orc.compression.strategy=COMPRESSION;
--分桶
--set hive.enforce.bucketing=true;
--set hive.enforce.sorting=true;
--set hive.optimize.bucketmapjoin = true;
--set hive.auto.convert.sortmerge.join=true;
--set hive.auto.convert.sortmerge.join.noconditionaltask=true;
--并行执行
set hive.exec.parallel=true;
set hive.exec.parallel.thread.number=8;
--小文件合并
-- set mapred.max.split.size=2147483648;
-- set mapred.min.split.size.per.node=1000000000;
-- set mapred.min.split.size.per.rack=1000000000;
--矢量化查询
set hive.vectorized.execution.enabled=true;
--关联优化器
set hive.optimize.correlation=true;
--读取零拷贝
set hive.exec.orc.zerocopy=true;
--join数据倾斜
set hive.optimize.skewjoin=false;
-- set hive.skewjoin.key=100000;
set hive.optimize.skewjoin.compiletime=false;
set hive.optimize.union.remove=false;
-- group倾斜
set hive.groupby.skewindata=false;
insert into table itcast_dws.class_attendance_dws partition(yearinfo,monthinfo,dayinfo)
select
concat(yearinfo,'-',monthinfo) as dateinfo,
class_id,
sum(studying_student_count) as studying_student_count,
sum(morning_att_count) as morning_att_count,
concat(round(sum(morning_att_count) / sum(studying_student_count) *100,2),'%') as morning_att_ratio,
sum(afternoon_att_count) as afternoon_att_count,
concat(round(sum(afternoon_att_count) / sum(studying_student_count) *100,2),'%') as afternoon_att_ratio,
sum(evening_att_count) as evening_att_count,
concat(round(sum(evening_att_count) / sum(studying_student_count) *100,2),'%') as evening_att_ratio,
sum(morning_late_count) as morning_late_count,
concat(round(sum(morning_late_count) / sum(studying_student_count) *100,2),'%') as morning_late_ratio,
sum(afternoon_late_count) as afternoon_late_count,
concat(round(sum(afternoon_late_count) / sum(studying_student_count) *100,2),'%') as afternoon_late_ratio,
sum(evening_late_count) as evening_late_count,
concat(round(sum(evening_late_count) / sum(studying_student_count) *100,2),'%') as evening_late_ratio,
sum(morning_leave_count) as morning_leave_count,
concat(round(sum(morning_leave_count) / sum(studying_student_count) *100,2),'%') as morning_leave_ratio,
sum(afternoon_leave_count) as afternoon_leave_count,
concat(round(sum(afternoon_leave_count) / sum(studying_student_count) *100,2),'%') as afternoon_leave_ratio,
sum(evening_leave_count) as evening_leave_count,
concat(round(sum(evening_leave_count) / sum(studying_student_count) *100,2),'%') as evening_leave_ratio,
sum(morning_truant_count) as morning_truant_count,
concat(round(sum(morning_truant_count) / sum(studying_student_count) *100,2),'%') as morning_truant_ratio,
sum(afternoon_truant_count) as afternoon_truant_count,
concat(round(sum(afternoon_truant_count) / sum(studying_student_count) *100,2),'%') as afternoon_truant_ratio,
sum(evening_truant_count) as evening_truant_count,
concat(round(sum(evening_truant_count) / sum(studying_student_count) *100,2),'%') as evening_truant_ratio,
'4' as time_type,
yearinfo,
monthinfo,
'-1' as dayinfo
from itcast_dwm.class_all_dwm
group by yearinfo,monthinfo,class_id;