12. 离线处理之业务数据采集、生成用户画像、推广效果分析以及知识点总结

数据库 2020-07-23 11:04:11 阅读次数: 0

离线处理之业务数据采集、生成用户画像、推广效果分析以及知识点总结

1. Azkaban周期性调度任务

1.1. 总览

在这里插入图片描述
在这里插入图片描述
在这里插入图片描述

1.2. 调度脚本

  • system_pre.properties

    dbUrl=jdbc:postgresql://rm-2zeqbua7952ni0c14.pg.rds.aliyuncs.com:3433/medchat
userName=******
password=******

  • bass_get_data.job

    type=command
command=sh sqoop_coll.sh ${dbUrl} ${userName} ${password}

  • bass_serv_profile.job

    type=command
command=sh bass_serv_profile.sh
dependencies=bass_get_data

  • bass_serv_profile.sh

    #!/bin/bash
echo "run profile start `date '+%Y-%m-%d,%H:%m:%s'`"
lastDay=`date -d "yesterday" +%Y-%m-%d`
#生成用户画像
hive -hiveconf lastDay=$lastDay -f serv_profile.sql
echo "run profile end `date '+%Y-%m-%d,%H:%m:%s'`"

  • bass_ad_analysis.job

    type=command
command=sh bass_ad_analysis.sh ${dbUrl} ${userName} ${password}
dependencies=bass_serv_profile

  • bass_ad_analysis.sh

    #!/bin/bash
echo "run ad analysis start `date '+%Y-%m-%d,%H:%m:%s'`"
dbUrl=$1
userName=$2
password=$3
lastDay=`date -d "yesterday" +%Y-%m-%d`
#广告效果分析
hive -hiveconf lastDay=$lastDay -hiveconf dbUrl=$dbUrl -hiveconf userName=$userName -hiveconf password=$password -f ad_analysis.sql
echo "run analysis end `date '+%Y-%m-%d,%H:%m:%s'`"

1.3. Azkaban安装并设置定时任务Schedule以及邮件发送接收

2. 业务数据采集

2.1. 后台通过logback把业务接口日志写入到本地文件

2.1.1. logback配置文件

在这里插入图片描述

2.1.2. 拦截器当中记录接口日志

在这里插入图片描述
在这里插入图片描述

2.1.3. 本地日志目录

在这里插入图片描述

2.1.4. 日志格式

{
    "time":"2020-07-17 09:06:06.897",
    "modelName":"MINI-PORTAL",
    "host":"172.17.176.152",
    "thread":"http-nio-8902-exec-62",
    "level":"INFO ",
    "file":"MedChatLogger.java:110",
    "source":"adFile",
    "adLogonName":"",
    "os":"",
    "operFlag":true,
    "inputParam":{
        "logonName":[
            "15165428830"
        ],
        "logonPwd":[
            "******"
        ]
    },
    "channel":"ANDROID",
    "language":"en",
    "isNeedSync":false,
    "remote":"114.247.227.197",
    "message":"User login successfully",
    "version":"1.3.0-preview",
    "mac":"A1000037DFFE57",
    "url":"/html/gateway/api.ajax",
    "token":"",
    "result":[
        {
            "token":"49f1428d2aea459fb795799b8f841254",
            "isCanPublish":true,
            "isCanAudit":false,
            "isBelongEditorialBoard":false,
            "isBelongTeam":true,
            "isHaveEducation":false,
            "isHaveWorkExperience":false,
            "authFlag":"Y",
            "isForcedAuth":false,
            "servId":"100034",
            "imPwd":"5ED47664D3C844ED075D121089B10DD8",
            "pushId":"100034",
            "pushAlias":"MOBILE",
            "servName":"文飞扬",
            "servIcon":"doctor/headportrait/person/1566811863179164958.jpg",
            "gender":"",
            "servInfoDegree":"1",
            "isVoice":false,
            "isShock":false,
            "isCertApply":false,
            "isHaveChannel":true,
            "cityName":"Yanbian Korean Autonomous Prefecture",
            "isSpecialServ":false,
            "isShowZone":true,
            "isHaveAuth":true,
            "isHaveProfile":false,
            "isCanNotice":false,
            "dutyDesc":"",
            "channelEduPublishType":"SPEECH",
            "mainDesc":"北京医院",
            "noticeInfo":{
                "noticeChatId":null,
                "noticeChatName":null
            },
            "isSwitchPhone":false,
            "isShowMobile":true,
            "isOpenEduInvite":false,
            "isAnonymous":false,
            "paymentSalt":"MTUxNjU1Mjg4NjQ=",
            "isRealName":false,
            "isHavePaymentPassword":false,
            "maxBankCardCount":"5",
            "menuArr":[
                "ARTICLE",
                "PEER",
                "CHAT",
                "ME"
            ],
            "servIdentityType":"DOCTOR",
            "isRemindDisturb":false,
            "eduCount":"1",
            "valueAddedService":"0"
        }
    ],
    "subErrcode":"",
    "carrier":"",
    "GATEWAY_MEDCHAT":"GATEWAY_MEDCHAT",
    "errCode":"",
    "model":"",
    "adServId":"0",
    "net":"",
    "apiType":"LOGIN"
}

{
    "time":"2020-07-17 10:26:59.535",
    "modelName":"MINI-PORTAL",
    "host":"172.17.176.152",
    "thread":"http-nio-8902-exec-82",
    "level":"INFO ",
    "file":"MedChatLogger.java:110",
    "source":"adFile",
    "apiType":"QUERY_CEDU_PLATFORM_INFO_LIST@AD_INFO",
    "adServId":"100242",
    "result":[
        {
            "adId":"657"
        }
    ]
}

2.2. 通过Flume采集数据到Kafka

2.2.1. Flume配置文件

# Each channel's type is defined.
#agent.channels.memoryChannel.type = memory

# Other config values specific to each type of channel(sink or source)
# can be defined as well
# In this case, it specifies the capacity of the memory channel
#agent.channels.memoryChannel.capacity = 100
# Name the components on this agent
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# Describe/configure the source
#a1.sources.r1.type = exec
#a1.sources.r1.command = tail -F /opt/log/ad-mini-portal3.log
a1.sources.r1.type = TAILDIR
a1.sources.r1.positionFile = /home/medchat/apache-flume-1.8.0-bin/logfile_stats/taildir_position.json
a1.sources.r1.filegroups = f1 f2 f3
a1.sources.r1.filegroups.f1 = /home/medchat/medchat-mini-portal/apache-tomcat/logs/doctor/ad-mini-portal.*log
a1.sources.r1.filegroups.f2 = /home/medchat/medchat-console/apache-tomcat-8.0.21/logs/console/ad-console.*log
a1.sources.r1.filegroups.f3 = /home/medchat/medchat-portal/apache-tomcat-8.0.21/logs/portal-yxck/ad-portal.*log
a1.sources.r1.fileHeader = true

# Describe the sink
a1.sinks.k1.type = org.apache.flume.sink.kafka.KafkaSink
a1.sinks.k1.kafka.bootstrap.servers= 172.17.176.160:9092,172.17.176.159:9092,172.17.176.158:9092
a1.sinks.k1.kafka.topic= adMiniPortal
a1.sinks.k1.serializer.class=kafka.serializer.StringEncoder
a1.sinks.k1.kafka.producer.acks=1
a1.sinks.k1.custom.encoding=UTF-8


# Use a channel which buffers events in memory
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# Bind the source and sink to the channel
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

在这里插入图片描述

2.3. Storm消费Kafka数据，写入Hbase

2.3.1. LogReaderSpout.java

package com.nuhtech.marketing.rta.storm.spouts;

import java.time.Duration;
import java.util.Map;

import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.apache.kafka.clients.consumer.KafkaConsumer;
import org.apache.storm.spout.SpoutOutputCollector;
import org.apache.storm.task.TopologyContext;
import org.apache.storm.topology.OutputFieldsDeclarer;
import org.apache.storm.topology.base.BaseRichSpout;
import org.apache.storm.tuple.Fields;
import org.apache.storm.tuple.Values;

import com.nuhtech.marketing.rta.util.KafkaUtil;
import com.nuhtech.medchat.core.util.PropConfigUtil;


public class LogReaderSpout extends BaseRichSpout {

    private static final String TOPIC = PropConfigUtil.getProperty("kafka.topic");

    private transient KafkaConsumer<String, String> consumer;

    private transient SpoutOutputCollector collector;

    @Override
    public void open(Map conf, TopologyContext context, SpoutOutputCollector collector) {
        this.collector = collector;
        this.consumer = KafkaUtil.getConsumer(TOPIC);
    }

    @Override
    public void nextTuple() {
        ConsumerRecords<String, String> messageList = consumer.poll(Duration.ofSeconds(4));
        messageList.forEach(message -> {
            if (message.value() != null && !"".equals(message.value())) {
                this.collector.emit(new Values(message.value()));
            }
        });

    }

    @Override
    public void declareOutputFields(OutputFieldsDeclarer declarer) {
        declarer.declare(new Fields("value"));
    }
}

2.3.2. UserLogBolt.java

package com.nuhtech.marketing.rta.storm.bolts;

import java.io.IOException;
import java.sql.Timestamp;
import java.util.HashMap;
import java.util.Map;

import net.sf.json.JSONArray;
import net.sf.json.JSONObject;

import org.apache.storm.topology.BasicOutputCollector;
import org.apache.storm.topology.OutputFieldsDeclarer;
import org.apache.storm.topology.base.BaseBasicBolt;
import org.apache.storm.tuple.Tuple;

import com.nuhtech.marketing.core.logger.MarketingLogger;
import com.nuhtech.marketing.core.util.HbaseUtil;
import com.nuhtech.medchat.core.util.DatetimeUtil;

public class UserLogBolt extends BaseBasicBolt {

    private static final long serialVersionUID = -5627408245880685574L;

    private static String tableName = "user_log";

    @Override
    public void execute(Tuple input, BasicOutputCollector collector) {
        String log = input.getStringByField("value");
        try {
            JSONObject jsonObject = JSONObject.fromObject(log);
            recordAll(log, jsonObject);
            recordAds(jsonObject);
        } catch (Exception e) {
            MarketingLogger.error(e);
        }
    }

    //记录所有行为日志
    public void recordAll(String log, JSONObject jsonObject) throws IOException {
        Timestamp timeStamp = DatetimeUtil.string2Timestamp(jsonObject.getString("time"));
        String adServId = jsonObject.getString("adServId");
        String time = DatetimeUtil.date2StringDateTimeNoLine(timeStamp);
        int id = (int) ((Math.random() * 9 + 1) * 10000);
        String servId = String.format("%08d", Long.parseLong(adServId));
        //rowKey:14位时间戳+8位servId+5位随机Id
        String rowKey = time + "|" + servId + "|" + id;
        Map<String, Object> map = new HashMap<String, Object>();
        map.put("rowKey", rowKey);
        map.put("columnFamily", "action");
        map.put("columnName", "info");
        map.put("columnValue", log);
        HbaseUtil.insertOnly(tableName, map);
    }

    //广告统计相关
    public void recordAds(JSONObject jsonObject) throws IOException {
        String adServId = jsonObject.getString("adServId");
        String apiType = jsonObject.getString("apiType");
        if (apiType.indexOf("@AD_INFO") != -1) {
            JSONArray jsonArray = jsonObject.getJSONArray("result");
            String ids = jsonArray.getJSONObject(0).getString("adId");
            String[] idsArr = ids.split(",");
            for (String id : idsArr) {
                String rowKey = adServId + "|" + id;
                HbaseUtil.autoIncrementColumn("realtime_ad_stat", rowKey, "show_cnt", 1);
            }
        }
    }

    @Override
    public void declareOutputFields(OutputFieldsDeclarer declarer) {

    }
}

2.3.3 ToHbaseBolt.java

package com.nuhtech.marketing.rta.storm.bolts;

import java.io.IOException;
import java.util.Map;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.hbase.HBaseConfiguration;
import org.apache.hadoop.hbase.TableName;
import org.apache.hadoop.hbase.client.Connection;
import org.apache.hadoop.hbase.client.ConnectionFactory;
import org.apache.hadoop.hbase.client.Table;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.storm.task.TopologyContext;
import org.apache.storm.topology.BasicOutputCollector;
import org.apache.storm.topology.OutputFieldsDeclarer;
import org.apache.storm.topology.base.BaseBasicBolt;
import org.apache.storm.tuple.Tuple;

import com.nuhtech.marketing.core.logger.MarketingLogger;

public class ToHbaseBolt extends BaseBasicBolt {

    private static final long serialVersionUID = -5627408245880685574L;

    private transient Table table;

    @SuppressWarnings("rawtypes")
    @Override
    public void prepare(Map stormConf, TopologyContext context) {
        Configuration conf = HBaseConfiguration.create();
        conf.set("hbase.zookeeper.quorum", "192.168.1.23:2181,192.168.1.24:2181,192.168.1.25:2181");
        try (Connection conn = ConnectionFactory.createConnection(conf)) {
            table = conn.getTable(TableName.valueOf("realtime_ad_stat"));
        } catch (IOException e) {
            MarketingLogger.error(e);
        }
    }

    @Override
    public void execute(Tuple input, BasicOutputCollector collector) {
        String action = input.getStringByField("action");
        String rowkey = input.getStringByField("rowkey");
        Long pv = input.getLongByField("cnt");

        try {
            if ("view".equals(action)) {
                table.incrementColumnValue(Bytes.toBytes(rowkey), Bytes.toBytes("stat"), Bytes.toBytes("view_cnt"), pv);
            }
            if ("click".equals(action)) {
                table.incrementColumnValue(Bytes.toBytes(rowkey), Bytes.toBytes("stat"), Bytes.toBytes("click_cnt"), pv);
            }
        } catch (IOException e) {
            MarketingLogger.error(e);
        }
    }

    @Override
    public void declareOutputFields(OutputFieldsDeclarer declarer) {
        // Do nothing
    }
}

2.4. Hive和Hbase的整合

2.4.1. HIve和Hbase的表关联

在这里插入图片描述

2.4.2. Hive数据样式

在这里插入图片描述

2.4.3. Hbase数据样式

在这里插入图片描述

2.4.4. Hive和Hbase关联后，操作一方同样会影响另一方数据

2.5. 通过Sqoop把业务数据从PostgreSql导入Hive数仓

#!/bin/bash
echo "sqoop coll start `date '+%Y-%m-%d,%H:%m:%s'`"
if [ $# -eq 3 ] ;then
echo "the args you input is right"
CONNECTURL=$1
USERNAME=$2
PASSWORD=$3
lastDay=`date -d "yesterday" +%Y-%m-%d`
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table f_serv --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table f_serv --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table f_serv_identity --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table f_serv_identity --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table f_serv_base --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table f_serv_base --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table s_hospital --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table s_hospital --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table f_channel_subscribe --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table f_channel_subscribe --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table mr_editorial_board_member --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table mr_editorial_board_member --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table e_presidium_member --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table e_presidium_member --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table e_rostrum_member --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table e_rostrum_member --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table e_edu_member --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table e_edu_member --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table mr_article_read_record --columns "read_record_id, article_id, serv_id, read_start_time, read_end_time, duration, update_datetime, create_datetime, thesis_type, spread_plan_id" --where "create_datetime::date='${lastDay}'" --hive-partition-key "dt" --hive-partition-value "${lastDay}" --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table mr_article_read_record --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table mr_article_laud --columns "article_laud_id, article_id, serv_id,update_datetime, create_datetime, thesis_type, spread_plan_id" --where "create_datetime::date='${lastDay}'" --hive-partition-key "dt" --hive-partition-value "${lastDay}" --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table mr_article_laud --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table mr_article_comment --columns "article_comment_id,article_id,serv_id,comment_type,ref_article_comment_id,content,ref_serv_id,status,update_datetime,create_datetime,content_original,thesis_type,spread_plan_id" --where "create_datetime::date='${lastDay}'" --hive-partition-key "dt" --hive-partition-value "${lastDay}" --null-string '\\N' --null-non-string '\\N' --fields-terminated-by '\001' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table mr_article_comment --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table mr_forward --columns "forward_id,serv_id,forward_type,forward_channel,refer_id,status,update_datetime,create_datetime,content,source_serv_id,target_serv_ids,target_group_id" --where "create_datetime::date='${lastDay}'" --hive-partition-key "dt" --hive-partition-value "${lastDay}" --null-string '\\N' --null-non-string '\\N' --fields-terminated-by '\001' --hive-drop-import-delims --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table mr_forward --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table mr_favorite --columns "favorite_id,serv_id,favorite_type,type,favorite_json,favorite_serv_id,refer_id,content,duration,favorite_url,status,update_datetime,create_datetime,file_size,refer_type" --where "create_datetime::date='${lastDay}'" --hive-partition-key "dt" --hive-partition-value "${lastDay}" --null-string '\\N' --null-non-string '\\N' --fields-terminated-by '\001' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table mr_favorite --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table sp_spread_plan --columns "spread_plan_id,service_provider_id,scheme_id,audit_status,publish_status,release_date,exp_date,update_datetime,create_datetime" --null-string '\\N' --null-non-string '\\N' --fields-terminated-by ',' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table sp_spread_plan --hive-overwrite
sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --table sp_plan_filter --columns "seq_nbr, spread_plan_id, serv_id, update_datetime, create_datetime" --where "create_datetime::date='${lastDay}'" --hive-partition-key "dt" --hive-partition-value "${lastDay}" --null-string '\\N' --null-non-string '\\N' --fields-terminated-by '\001' --delete-target-dir --num-mappers 1 --hive-import --hive-database default --hive-table sp_plan_filter --hive-overwrite

hadoop fs -test -e /user/hive/warehouse/mr_plan_article
if [ $? -ne 0 ]; then
    sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --query "select a.article_id,a.sale_article_id,b.spread_plan_id from mr_article a,sp_sale_article b where a.dispatch_type='SALE' and a.sale_article_id=b.sale_article_id and b.spread_plan_id is not null and \$CONDITIONS" --fields-terminated-by '\001' --m 1 --lines-terminated-by '\n' --hive-drop-import-delims --split-by article_id --target-dir /user/hive/warehouse/mr_plan_article
else
    hdfs dfs -rm -r /user/hive/warehouse/mr_plan_article
    sqoop import --connect $CONNECTURL --username $USERNAME --password $PASSWORD --query "select a.article_id,a.sale_article_id,b.spread_plan_id from mr_article a,sp_sale_article b where a.dispatch_type='SALE' and a.sale_article_id=b.sale_article_id and b.spread_plan_id is not null and \$CONDITIONS" --fields-terminated-by '\001' --m 1 --lines-terminated-by '\n' --hive-drop-import-delims --split-by article_id --target-dir /user/hive/warehouse/mr_plan_article
fi
else
echo "the args you input is error"
fi
echo "sqoop coll end `date '+%Y-%m-%d,%H:%m:%s'`"

  • f_serv用户表
    在这里插入图片描述

  • f_serv_identity用户身份信息表
    在这里插入图片描述

  • f_serv_base用户基本信息表
    在这里插入图片描述

  • s_hospital医院表
    在这里插入图片描述

  • f_channel_subscribe学科订阅表
    在这里插入图片描述

  • mr_editorial_board_member学术组织成员表
    在这里插入图片描述

  • e_presidium_member主席团成员表
    在这里插入图片描述

  • e_rostrum_member主席台成员表
    在这里插入图片描述

  • e_edu_member会场成员表
    在这里插入图片描述

  • mr_article_read_record文章阅读表
    在这里插入图片描述

  • mr_article_laud文章点赞表
    在这里插入图片描述

  • mr_article_comment文章评论表
    在这里插入图片描述

  • mr_forward用户转发表
    在这里插入图片描述
    在这里插入图片描述

  • mr_favorite用户收藏表
    在这里插入图片描述

  • sp_spread_plan推广计划表
    在这里插入图片描述

  • sp_plan_filter广告过滤表
    在这里插入图片描述

  • 注意点

3. 生成用户画像

3.1. hive分区设置

set hive.exec.dynamic.partition=true;
set hive.exec.dynamic.partition.mode=nonstrict;

3.2. 用户登录日志

insert overwrite table s_login_log partition(dt)    -- 分区
select
       get_json_object(info, '$.result[0].servId'),
       get_json_object(info, '$.modelName'),    -- 后台模型名称
       get_json_object(info, '$.channel') ,     -- terminal_type 终端类型
       get_json_object(info, '$.model'),        -- phone_model 手机机型
       get_json_object(info, '$.time'),         -- create_datetime 创建时间
       substr(get_json_object(info, '$.time'),1,10) -- 分区字段
from user_log
where
      get_json_object(info, '$.apiType')='LOGIN' and get_json_object(info, '$.operFlag')='true' and substr(get_json_object(info, '$.time'),1,10)='${hiveconf:lastDay}';

在这里插入图片描述

3.3. 用户终端类型和机型

insert into serv_terminal
select
       serv_id,terminal_type,phone_model
from
    (select
            serv_id,terminal_type,phone_model,max(create_datetime)
    from s_login_log
    where dt='${hiveconf:lastDay}'
    group by
             serv_id,terminal_type,phone_model
    ) a;

在这里插入图片描述

3.4. 用户最高委员会职务

insert overwrite table mr_newspaper_duty_highest
select
       serv_id,newspaper_title,newspaper_title_order
from
     (select row_number() over (partition by serv_id order by newspaper_title_order) top,*
     from
        (select t.serv_id,n.newspaper_title,n.newspaper_title_name,newspaper_title_order
        from
             mr_editorial_board_member t,mr_newspaper_title n
        where
              t.newspaper_title = n.newspaper_title
        )as b
    ) as c
where
      top<=1
order by serv_id,newspaper_title_order;

在这里插入图片描述

3.5. 用户信息

insert overwrite table f_serv_info
select
       a.serv_id,a.identity,g.terminal_type,
       case when length(b.province_code)!=6 then null else b.province_code end  province_code,b.city_code,
       case when b.gender in('F','M') then b.gender else null end gender,
       case when a.identity='DOCTOR' and c.hospital_duty is not null then c.hospital_duty else 'DOC_QT' end hospital_duty,
       case when a.identity='DOCTOR' and c.hospital_title is not null then c.hospital_title else 'OTH' end hospital_title,
       case when a.identity='DOCTOR' then c.department_id else null end department_id,
       case when a.identity='DOCTOR'  and f.department_type_id is not null then f.department_type_id  else 27 end department_type_id,
       case when a.identity='DOCTOR' then d.hospital_id else null end hospital_id,
       case when a.identity='DOCTOR' and d.hospital_level!='' then d.hospital_level else null end hospital_level,
       e.newspaper_title
from f_serv a
left join f_serv_base b on(a.serv_id=b.serv_id)
left join f_serv_identity c on(a.serv_id=c.serv_id)
left join s_hospital d on(c.hospital_id=d.hospital_id)
left join mr_newspaper_duty_highest e on(a.serv_id=e.serv_id)
left join s_department f on(c.department_id=f.department_id)
left join serv_terminal g on(a.serv_id=g.serv_id)
where a.auth_flag in('Y','R');

在这里插入图片描述

3.6. 画像信息

  • 个人信息标签

    insert overwrite table f_serv_tag_value  partition(module='basic_info')
select
       serv_id, mp['key'], mp['value']
from(
    select a.serv_id,
             array(map('key', 'identity', 'value', a.identity),
                    map('key', 'device_type', 'value', a.terminal_type),
                    map('key', 'gender', 'value',a.gender),
                    map('key', 'province_code', 'value', a.province_code),
                    map('key', 'city_code', 'value', a.city_code),
                    map('key', 'hospital_id', 'value', a.hospital_id),
                    map('key', 'department_id', 'value', a.department_id),
                    map('key', 'department_type_id', 'value', a.department_type_id),
                    map('key', 'hospital_title', 'value', a.hospital_title),
                    map('key', 'hospital_duty', 'value', a.hospital_duty),
                    map('key', 'hospital_level', 'value', a.hospital_level)
                   ) arr
    from f_serv_info a
    ) s  lateral view explode(arr) arrtable as mp;

    在这里插入图片描述

  • 频道标签

    insert overwrite table f_serv_tag_value  partition(module='channel_subscribe')
select
       serv_id, mp['key'], mp['value']
from
     (select a.serv_id,array(map('key', 'channel_subscribe', 'value', a.channel_list)) arr
     from
          (select
               serv_id,concat_ws(',',collect_list(cast(channel_id as string))) as channel_list
          from
               f_channel_subscribe
          group by
               serv_id
          ) a
    )s  lateral view explode(arr) arrtable as mp;

    在这里插入图片描述

  • 委员会标签

    insert overwrite table f_serv_tag_value  partition(module='newspaper_title')
select serv_id, mp['key'], mp['value']
from(
select a.serv_id,array(map('key', 'newspaper_title', 'value', a.newspaper_title)) arr
from (select serv_id,concat_ws(',',collect_set(newspaper_title)) newspaper_title from mr_editorial_board_member group by serv_id) a
)s  lateral view explode(arr) arrtable as mp;

    在这里插入图片描述

  • 演讲台标签

    insert overwrite table f_serv_tag_value  partition(module='edu')
select serv_id, mp['key'], mp['value']
from(
select b.serv_id,array(map('key', 'channel_edu', 'value', b.channel_edu_id)) arr
from (select serv_id,concat_ws(',',collect_set(cast(channel_edu_id as string))) channel_edu_id
from (select channel_edu_id,serv_id from e_presidium_member
union
select channel_edu_id,serv_id from e_rostrum_member
union
select channel_edu_id,serv_id from e_edu_member)  a group by serv_id) b
)s  lateral view explode(arr) arrtable as mp;

    在这里插入图片描述

3.7. 生成用户画像

insert overwrite table f_serv_profile
select
    serv_id,
    concat('{', concat_ws(',', collect_set(concat('"', tag, '"', ':', '"', value, '"'))), '}') as json_string
from f_serv_tag_value where serv_id is not null
group by serv_id ;

insert into table serv_sysc select serv_id,profile from f_serv_profile ;

在这里插入图片描述

4. 推广效果分析

4.1. 采集微论文曝光数据

insert into table mr_article_show_record  partition(dt='${hiveconf:lastDay}')
select
       get_json_object(info, '$.apiType') apiType,
       num adId,
       get_json_object(info, '$.adServId') servId,
       get_json_object(info, '$.time') time
from user_log LATERAL VIEW explode(split(get_json_object(info, '$.result[0].adId'),',')) zqm AS num
where
      get_json_object(info, '$.apiType') like '%@ARTICLE_INFO' and substr(get_json_object(info, '$.time'),1,10)='${hiveconf:lastDay}';

在这里插入图片描述

4.2. 采集广告曝光数据

insert overwrite table mr_ad_show_record  partition(dt='${hiveconf:lastDay}')
select
       get_json_object(info, '$.apiType') apiType,
       num adId,
       get_json_object(info, '$.adServId') servId,
       get_json_object(info, '$.time') time
from user_log LATERAL VIEW explode(split(get_json_object(info, '$.result[0].adId'),',')) zqm AS num
where
      get_json_object(info, '$.apiType') like '%@AD_INFO' and substr(get_json_object(info, '$.time'),1,10)='${hiveconf:lastDay}';

在这里插入图片描述

  • 合并微文论曝光数据
    insert into table mr_ad_show_record  partition(dt='${hiveconf:lastDay}')
select
       a.api_type,
       b.spread_plan_id,
       a.serv_id,
       a.show_time
from mr_article_show_record a,mr_plan_article b where a.article_id=b.article_id and a.dt='${hiveconf:lastDay}'

4.3. 曝光量多维度统计

insert overwrite table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,
       b.identity,
       b.terminal_type,
       b.province_code,
       b.gender,
       b.hospital_duty,
       b.hospital_title,
       b.department_type_id,
       b.hospital_level,
       b.newspaper_title,
       'show' analysis_type,
       GROUPING__ID,
       count(a.serv_id) active_num
from mr_ad_show_record a
join f_serv_info b on(a.serv_id=b.serv_id)
where a.dt='${hiveconf:lastDay}'
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

-- 点击量多维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'click' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select spread_plan_id,serv_id from mr_article_read_record where thesis_type='ADVERT' and dt='${hiveconf:lastDay}'
    union all
    select p.spread_plan_id,t.serv_id from mr_article_read_record t,mr_plan_article p where t.article_id=p.article_id and t.thesis_type='ARTICLE' and t.dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

--阅读量多维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'read' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select spread_plan_id,serv_id from mr_article_read_record where thesis_type='ADVERT' and duration>=2 and dt='${hiveconf:lastDay}'
    union all
    select p.spread_plan_id,t.serv_id from mr_article_read_record t,mr_plan_article p where t.article_id=p.article_id and t.thesis_type='ARTICLE' and t.duration>=2 and t.dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

--点赞量多维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'up' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select spread_plan_id,serv_id from mr_article_laud where thesis_type='ADVERT' and dt='${hiveconf:lastDay}'
    union all
    select p.spread_plan_id,t.serv_id from mr_article_laud t,mr_plan_article p where t.article_id=p.article_id and t.thesis_type='ARTICLE' and t.dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

--评论量多维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'comment' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select spread_plan_id,serv_id from mr_article_comment where thesis_type='ADVERT' and dt='${hiveconf:lastDay}'
    union all
    select p.spread_plan_id,t.serv_id from mr_article_comment t,mr_plan_article p where t.article_id=p.article_id and t.thesis_type='ARTICLE' and t.dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

--收藏量多维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'coll' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select refer_id spread_plan_id,serv_id from mr_favorite where refer_type='ADVERT' and dt='${hiveconf:lastDay}'
    union all
    select p.spread_plan_id,t.serv_id from mr_favorite t,mr_plan_article p where t.refer_id=p.article_id and t.refer_type='ARTICLE' and t.dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

--转发量多维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'forward' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select refer_id spread_plan_id,serv_id from mr_forward where forward_type='20' and dt='${hiveconf:lastDay}'
    union all
    select p.spread_plan_id,t.serv_id from mr_forward t,mr_plan_article p where t.refer_id=p.article_id and t.forward_type='1' and  t.dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

--问卷多维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'answer' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select spread_plan_id,serv_id from sp_serv_reply where dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

--用户不敢兴趣维度统计
insert into table rp_spread_analysis_detail partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title,'filter' analysis_type,GROUPING__ID,count(a.serv_id) active_num
from
    (select spread_plan_id,serv_id from sp_plan_filter where dt='${hiveconf:lastDay}') a
join f_serv_info b on(a.serv_id=b.serv_id)
group by  a.spread_plan_id,b.identity,b.terminal_type,b.province_code,b.gender,b.hospital_duty,b.hospital_title,b.department_type_id,b.hospital_level,b.newspaper_title
grouping sets ((a.spread_plan_id,b.identity),(a.spread_plan_id,b.terminal_type),(a.spread_plan_id,b.province_code),(a.spread_plan_id,b.gender),(a.spread_plan_id,b.hospital_duty),(a.spread_plan_id,b.hospital_title),(a.spread_plan_id,b.department_type_id),(a.spread_plan_id,b.hospital_level),(a.spread_plan_id,b.newspaper_title));

在这里插入图片描述

4.4. 不同维度进行曝光量的统计

--统计身份维度
insert overwrite table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'identity' attr_type,a.identity attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,identity,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='11111111'
    group by spread_plan_id,identity) a;

--统计医院等级维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'hospitalLevel' attr_type,a.hospital_level attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select spread_plan_id,hospital_level,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='111111101'
    group by spread_plan_id,hospital_level) a;

--统计职称维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'hospitalTitle' attr_type,a.hospital_title attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,hospital_title,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='111110111'
    group by spread_plan_id,hospital_title) a;

--统计职务维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'hospitalDuty' attr_type,a.hospital_duty attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,hospital_duty,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='111101111'
    group by spread_plan_id,hospital_duty) a;

--统计性别维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'sex' attr_type,a.gender attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,gender,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='111011111'
    group by spread_plan_id,gender) a;
--统计机型维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'os' attr_type,a.terminal_type attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,terminal_type,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='101111111'
    group by spread_plan_id,terminal_type) a;
--统计省份维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'provinceCode' attr_type,a.province_code attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,province_code,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='110111111'
    group by spread_plan_id,province_code) a;

--统计科室分类维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
    a.spread_plan_id,'department' attr_type,a.department_id attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,department_id,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='111111011'
    group by spread_plan_id,department_id) a;

--统计委员会职务维度
insert into table rp_spread_analysis partition(dt='${hiveconf:lastDay}')
select
       a.spread_plan_id,'boardTitle' attr_type,a.newspaper_title attr_value,a.show_num,a.click_num,0 click_rate,a.read_num,a.up_num,a.comment_num,a.forward_num ,a.coll_num,'${hiveconf:lastDay}' total_date,a.answer_num,a.filter_num
from (
    select
        spread_plan_id,newspaper_title,
        sum(if(analysis_type='show',active_num,0)) as show_num,
        sum(if(analysis_type='click',active_num,0)) as click_num,
        sum(if(analysis_type='read',active_num,0)) as read_num ,
        sum(if(analysis_type='up',active_num,0)) as up_num ,
        sum(if(analysis_type='comment',active_num,0)) as comment_num,
        sum(if(analysis_type='coll',active_num,0)) as coll_num,
        sum(if(analysis_type='forward',active_num,0)) as forward_num,
        sum(if(analysis_type='answer',active_num,0)) as answer_num,
        sum(if(analysis_type='filter',active_num,0)) as filter_num
    from rp_spread_analysis_detail
    where dt='${hiveconf:lastDay}' and  bin(grouping_id)='111111110'
    group by spread_plan_id,newspaper_title) a;

4.5. 统计日汇总维度

insert overwrite table rp_spread partition(dt='${hiveconf:lastDay}')
select
       spread_plan_id,sum(show_num),sum(click_num),0,sum(read_num),sum(up_num),sum(comment_num),sum(forward_num),sum(coll_num),total_date,sum(answer_num),sum(filter_num)
from rp_spread_analysis
where total_date='${hiveconf:lastDay}'and attr_type='identity'
group by spread_plan_id,total_date;

在这里插入图片描述

4.6. 同步统计结果

SELECT dboutput('${hiveconf:dbUrl}', '${hiveconf:userName}', '${hiveconf:password}', 'INSERT INTO rp_spread (spread_plan_id, show_num, click_num, read_num, up_num, comment_num, forward_num,coll_num,answer_num,filter_num, total_date,create_datetime) VALUES (?,?,?,?,?,?,?,?,?,?,?,?);', spread_plan_id, show_num, click_num, read_num, up_num, comment_num, forward_num,coll_num,answer_num,filter_num, total_date, current_timestamp)
FROM rp_spread where total_date='${hiveconf:lastDay}';

SELECT dboutput('${hiveconf:dbUrl}', '${hiveconf:userName}', '${hiveconf:password}', 'INSERT INTO rp_spreed_analysis (spread_plan_id, attr_type, attr_value, show_num, click_num, read_num, up_num, comment_num, forward_num, coll_num,answer_num,filter_num, total_date, create_datetime) VALUES (?,?,?,?,?,?,?,?,?,?,?,?,?,?);',spread_plan_id, attr_type, attr_value, show_num, click_num, read_num, up_num, comment_num, forward_num, coll_num,answer_num,filter_num, total_date, current_timestamp)
FROM rp_spread_analysis where total_date='${hiveconf:lastDay}';

SELECT dboutput('${hiveconf:dbUrl}', '${hiveconf:userName}', '${hiveconf:password}', 'INSERT INTO rp_analysis_log (log_type,log_info,create_datetime) VALUES (?,?,?);', 'item_analysis','success', current_timestamp);

在这里插入图片描述
在这里插入图片描述
在这里插入图片描述

5. 知识点总结

5.1. Tomcat

5.1.1. Tomcat使用详细教程

5.1.2. 用脚本实现windows与linux之间文件的传输

5.2. Hive Hql常用方法总结

5.2.1. ROW_NUMBER() OVER函数的基本用法

5.2.2. SQL语言-- SELECT CASE WHEN THEN

5.2.3. Hive列转行 (Lateral View + explode)详解

5.2.4. HiveSQL行转列lateral view explore()以及连接concat_ws()和列转行collect_list()&collect_set()区别的使用案例

5.2.5. Hive SQL grouping sets 用法

6. 寄语：知己知彼，不狂不馁，仔细地找准了自己生命的目标，板浆摇橹向人生茫茫之海努力划去。

猜你喜欢

转载自blog.csdn.net/fanjianhai/article/details/107387578
今日推荐
TIOBE 5 月榜单：Fortran “复活”进入 Top 10
GCC 14.1 发布
面壁智能发布 Eurux-8x22B 开源大模型 —— 堪称「理科状元」
开源日报 | 谷歌扶持鸿蒙上位；开源Rabbit R1；Docker加持的安卓手机；微软的焦虑和野心；海尔电器把开放平台关了
中国码农的“35岁魔咒”
蘭雅 CorelDRAW 插件 2024.5.1 国际劳动节版，免费下载
Arc Browser for Windows 1.0 正式 GA
90后程序员开发视频搬运软件、不到一年获利超 700 万，结局很刑！
周排行
基本数据类型封装类比较 Java源码解读(一) 8种基本类型对应的封装类型
JS实现无缝滚动上
深入解析HashMap原理（基于JDK1.8）
mysql的连接池
关于.htc
linux下的ubuntu12.04图形界面
【数论】好推不好记的扩展欧几里德
设备树详解
cscope + tags 简单设置
xml学习
每日归档
更多
2024-05-09(35)
2024-05-08(42)
2024-05-07(14)
2024-05-06(40)
2024-05-05(0)
2024-05-04(7)
2024-05-03(19)
2024-05-02(0)
2024-05-01(4)
2024-04-30(1)

代码天地 © 2018 codetd.com

境外服务器   最新电视剧2022