Skip to content

M
meta_base_code
Commit 44f76a08 authored by litaolemo's avatar litaolemo
Browse files
Options

update

parent 1cc20431
Hide whitespace changes
Inline Side-by-side
Showing with 613 additions and 611 deletions
task/search_strategy_d.py
View file @ 44f76a08
...@@ -82,25 +82,444 @@ for t in range(1, task_days): ...@@ -82,25 +82,444 @@ for t in range(1, task_days):
yesterday_str = (now + datetime.timedelta(days=-1)).strftime("%Y%m%d") yesterday_str = (now + datetime.timedelta(days=-1)).strftime("%Y%m%d")
one_week_age_str = (now + datetime.timedelta(days=-7)).strftime("%Y%m%d") one_week_age_str = (now + datetime.timedelta(days=-7)).strftime("%Y%m%d")
sql_dev_device_id = """ sql_dev_device_id = """
SELECT partition_date,device_id SELECT partition_date,device_id
FROM
(--找出user_id当天活跃的第一个设备id
SELECT user_id,partition_date,
if(size(device_list) > 0, device_list [ 0 ], '') AS device_id
FROM online.ml_user_updates
WHERE partition_date>='{yesterday_str}' AND partition_date<'{today_str}'
)t1
JOIN
( --医生账号
SELECT distinct user_id
FROM online.tl_hdfs_doctor_view
WHERE partition_date = '{yesterday_str}'
--马甲账号/模特用户
UNION ALL
SELECT user_id
FROM ml.ml_c_ct_ui_user_dimen_d
WHERE partition_day = '{yesterday_str}'
AND (is_puppet = 'true' or is_classifyuser = 'true')
UNION ALL
--公司内网覆盖用户
select distinct user_id
from dim.dim_device_user_staff
UNION ALL
--登陆过医生设备
SELECT distinct t1.user_id
FROM
(
SELECT user_id, v.device_id as device_id
FROM online.ml_user_history_detail
LATERAL VIEW EXPLODE(device_history_list) v AS device_id
WHERE partition_date = '{yesterday_str}'
)t1
JOIN
(
SELECT device_id
FROM online.ml_device_history_detail
WHERE partition_date = '{yesterday_str}'
AND is_login_doctor = '1'
)t2
ON t1.device_id = t2.device_id
)t2
on t1.user_id=t2.user_id
group by partition_date,device_id
""".format(yesterday_str=yesterday_str, today_str=today_str)
print(sql_dev_device_id)
dev_df = spark.sql(sql_dev_device_id)
dev_df_view = dev_df.createOrReplaceTempView("dev_view")
dev_df.cache()
dev_df.show(1)
sql_res = dev_df.collect()
for res in sql_res:
print(res)
print("-------------------------------")
sql_spam_pv_device_id = """
SELECT DISTINCT device_id
FROM ml.ml_d_ct_dv_devicespam_d --去除机构刷单设备，即作弊设备（浏览和曝光事件去除）
WHERE partition_day='{yesterday_str}'
UNION ALL
SELECT DISTINCT device_id
FROM dim.dim_device_user_staff
""".format(yesterday_str=yesterday_str)
print(sql_spam_pv_device_id)
spam_pv_df = spark.sql(sql_spam_pv_device_id)
spam_pv_df.createOrReplaceTempView("spam_pv")
spam_pv_df.show(1)
sql_res = spam_pv_df.collect()
spam_pv_df.cache()
for res in sql_res:
print(res)
print("-------------------------------")
sql = r"""
SELECT t3.partition_date as partition_date
,t3.device_os_type as device_os_type
,t3.active_type as active_type
,t3.channel as channel
,NVL(t3.search_pv,0) as pv
,NVL(t3.search_uv,0) as uv
,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.hexin_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_core_pv
,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.neirong_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_pv
FROM
(--昨天总搜索量
SELECT partition_date,active_type,device_os_type,channel,search_pv,search_uv
FROM
(
SELECT t1.partition_date,active_type,device_os_type,channel
,count(t1.cl_id) as search_pv
,count(distinct t1.cl_id) as search_uv
FROM
(
SELECT partition_date
,params['query'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action in ('do_search','search_result_click_search')
UNION ALL
SELECT partition_date,params['query'] as query,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action = 'on_click_card'
AND params['page_name']='search_home'
UNION ALL
SELECT partition_date
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action = 'on_click_card'
AND params['in_page_pos']='猜你喜欢'
AND params['tab_name']='精选'
AND params['card_type']='search_word'
--AND page_name='home' android的page_name为空
UNION ALL
SELECT partition_date
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action = 'on_click_card'
AND page_name='welfare_home'
AND params['card_type'] ='search_word'
AND params['in_page_pos']='大家都在搜'
UNION ALL
SELECT partition_date
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND int(split(app_version,'\\.')[1]) >= 27
AND action='on_click_card'
AND params['card_type']='highlight_word'
)t1
JOIN
(
SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type
FROM
(
SELECT
partition_date,m.device_id
,array(device_os_type ,'合计') as device_os_type
,array(case WHEN active_type = '4' THEN '老活'
WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type
,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel
FROM online.ml_device_day_active_status m
LEFT JOIN
(SELECT code,is_ai_channel,partition_day
FROM DIM.DIM_AI_CHANNEL_ZP_NEW
WHERE partition_day>= {yesterday_str}
AND partition_day < {today_str}) tmp
ON m.partition_date=tmp.partition_day AND first_channel_source_type=code
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND active_type in ('1','2','4')
AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
,'promotion_shike','promotion_julang_jl03','promotion_zuimei')
AND first_channel_source_type not like 'promotion\_jf\_%'
) mas
LATERAL VIEW explode(mas.channel) t2 AS channel
LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type
LATERAL VIEW explode(mas.active_type) t2 AS active_type
)t2
on t1.cl_id=t2.device_id AND t1.partition_date = t2.partition_date
LEFT JOIN spam_pv
on spam_pv.device_id=t1.cl_id
LEFT JOIN dev_view
on t1.partition_date=dev_view.partition_date and t1.cl_id=dev_view.device_id
WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='')
and (dev_view.device_id is null or dev_view.device_id ='')
GROUP BY t1.partition_date,t2.active_type,device_os_type,channel
)t
)t3
LEFT JOIN
(--昨天搜索结果页卡片点击pv
SELECT t1.partition_date,active_type,device_os_type,channel
,sum(hexin) as hexin_card_click_pv
,sum(neirong) as neirong_card_click_pv
FROM
(
SELECT NVL(t2.partition_date,t3.partition_date) as partition_date
,NVL(t2.cl_id,t3.cl_id) as cl_id
,NVL(t2.query,t3.query) as query
,NVL(t2.pv,0) as hexin
,NVL(t3.pv,0) as neirong
FROM
(--核心卡片点击
SELECT partition_date
,params['query'] as query
,cl_id
,count(1) as pv
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND ((action in ('search_result_click_recommend_item','search_result_welfare_click_item','search_result_hospital_click_item','search_result_doctor_click_item','on_click_doctor_card', 'on_click_hospital_card')
AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor'))
or (action = 'goto_welfare_detail' AND params [ 'from' ] = 'search_result_welfare_recommend')
or (action = 'on_click_card' AND params['card_content_type'] in ('service','hospital','doctor') AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor'))
or (action = 'on_click_button' AND params['button_name'] = 'check_plan' AND page_name = 'search_result_more'))
GROUP BY partition_date
,params['query']
,cl_id
)t2
FULL JOIN
(--内容卡片点击
SELECT partition_date
,params['query'] as query
,cl_id
,count(1) as pv
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND ((action in ('on_click_topic_card','on_click_diary_card','search_result_click_infomation_item')
AND page_name in ('search_result_more','search_result_diary','search_result_post'))
or (action = 'on_click_card' AND params['card_content_type'] in ('answer','diary') AND page_name in ('search_result_more','search_result_diary','search_result_question_answer')))
GROUP BY partition_date
,params['query']
,cl_id
)t3
on t3.partition_date=t2.partition_date
AND t3.query=t2.query
AND t3.cl_id=t2.cl_id
)t1
JOIN
(
SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type
FROM
(
SELECT
partition_date,m.device_id
,array(device_os_type ,'合计') as device_os_type
,array(case WHEN active_type = '4' THEN '老活'
WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type
,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel
FROM online.ml_device_day_active_status m
LEFT JOIN
(SELECT code,is_ai_channel,partition_day
FROM DIM.DIM_AI_CHANNEL_ZP_NEW
WHERE partition_day>= {yesterday_str}
AND partition_day < {today_str}) tmp
ON m.partition_date=tmp.partition_day AND first_channel_source_type=code
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND active_type in ('1','2','4')
AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
,'promotion_shike','promotion_julang_jl03','promotion_zuimei')
AND first_channel_source_type not like 'promotion\_jf\_%'
) mas
LATERAL VIEW explode(mas.channel) t2 AS channel
LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type
LATERAL VIEW explode(mas.active_type) t2 AS active_type
)dev
on t1.cl_id=dev.device_id and t1.partition_date = dev.partition_date
LEFT JOIN spam_pv
on spam_pv.device_id=t1.cl_id
LEFT JOIN dev_view
on t1.partition_date=dev_view.partition_date and t1.cl_id=dev_view.device_id
WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='')
and (dev_view.device_id is null or dev_view.device_id ='')
GROUP BY t1.partition_date,active_type,device_os_type,channel
)t4
on t3.partition_date=t4.partition_date and t3.active_type=t4.active_type and t3.device_os_type = t4.device_os_type AND t3.channel = t4.channel
""".format(today_str=today_str, yesterday_str=yesterday_str)
device_df = spark.sql(sql)
device_df.show(1, False)
sql_res = device_df.collect()
for res in sql_res:
print(res)
device_df.createOrReplaceTempView("data_table")
sql = r"""
SELECT t3.partition_date as partition_date
,t3.device_os_type as device_os_type
,t3.active_type as active_type
,t3.channel as channel
,NVL(t3.search_pv,0) as pv
,NVL(t3.search_uv,0) as uv
,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.hexin_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_core_pv
,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.neirong_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_pv
FROM
(--昨天总搜索量
SELECT partition_date,active_type,device_os_type,channel,search_pv,search_uv
FROM
(
SELECT t1.partition_date,active_type,device_os_type,channel
,count(t1.cl_id) as search_pv
,count(distinct t1.cl_id) as search_uv
FROM
(
SELECT partition_date
,params['query'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action in ('do_search','search_result_click_search')
UNION ALL
SELECT partition_date,params['query'] as query,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action = 'on_click_card'
AND params['page_name']='search_home'
UNION ALL
SELECT partition_date
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action = 'on_click_card'
AND params['in_page_pos']='猜你喜欢'
AND params['tab_name']='精选'
AND params['card_type']='search_word'
--AND page_name='home' android的page_name为空
UNION ALL
SELECT partition_date
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action = 'on_click_card'
AND page_name='welfare_home'
AND params['card_type'] ='search_word'
AND params['in_page_pos']='大家都在搜'
UNION ALL
SELECT partition_date
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND int(split(app_version,'\\.')[1]) >= 27
AND action='on_click_card'
AND params['card_type']='highlight_word'
)t1
JOIN
(
SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type
FROM
(
SELECT
partition_date,m.device_id
,array(device_os_type ,'合计') as device_os_type
,array(case WHEN active_type = '4' THEN '老活'
WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type
,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel
FROM online.ml_device_day_active_status m
LEFT JOIN
(SELECT code,is_ai_channel,partition_day
FROM DIM.DIM_AI_CHANNEL_ZP_NEW
WHERE partition_day>= {yesterday_str}
AND partition_day < {today_str}) tmp
ON m.partition_date=tmp.partition_day AND first_channel_source_type=code
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND active_type in ('1','2','4')
AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
,'promotion_shike','promotion_julang_jl03','promotion_zuimei')
AND first_channel_source_type not like 'promotion\_jf\_%'
) mas
LATERAL VIEW explode(mas.channel) t2 AS channel
LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type
LATERAL VIEW explode(mas.active_type) t2 AS active_type
)t2
on t1.cl_id=t2.device_id AND t1.partition_date = t2.partition_date
LEFT JOIN
(
SELECT DISTINCT device_id
FROM ml.ml_d_ct_dv_devicespam_d --去除机构刷单设备，即作弊设备（浏览和曝光事件去除）
WHERE partition_day={yesterday_str}
UNION ALL
SELECT DISTINCT device_id
FROM dim.dim_device_user_staff --去除内网用户
)spam_pv
on spam_pv.device_id=t1.cl_id
LEFT JOIN
(
SELECT partition_date,device_id
FROM FROM
(--找出user_id当天活跃的第一个设备id (--找出user_id当天活跃的第一个设备id
SELECT user_id,partition_date, SELECT user_id,partition_date,
if(size(device_list) > 0, device_list [ 0 ], '') AS device_id if(size(device_list) > 0, device_list [ 0 ], '') AS device_id
FROM online.ml_user_updates FROM online.ml_user_updates
WHERE partition_date>='{yesterday_str}' AND partition_date<'{today_str}' WHERE partition_date>='${{yesterday_str}}' AND partition_date<'${{today_str}}'
)t1 )t1
JOIN JOIN
( --医生账号 ( --医生账号
SELECT distinct user_id SELECT distinct user_id
FROM online.tl_hdfs_doctor_view FROM online.tl_hdfs_doctor_view
WHERE partition_date = '{yesterday_str}' WHERE partition_date = {yesterday_str}
--马甲账号/模特用户 --马甲账号/模特用户
UNION ALL UNION ALL
SELECT user_id SELECT user_id
FROM ml.ml_c_ct_ui_user_dimen_d FROM ml.ml_c_ct_ui_user_dimen_d
WHERE partition_day = '{yesterday_str}' WHERE partition_day = {yesterday_str}
AND (is_puppet = 'true' or is_classifyuser = 'true') AND (is_puppet = 'true' or is_classifyuser = 'true')
UNION ALL UNION ALL
...@@ -116,262 +535,183 @@ for t in range(1, task_days): ...@@ -116,262 +535,183 @@ for t in range(1, task_days):
SELECT user_id, v.device_id as device_id SELECT user_id, v.device_id as device_id
FROM online.ml_user_history_detail FROM online.ml_user_history_detail
LATERAL VIEW EXPLODE(device_history_list) v AS device_id LATERAL VIEW EXPLODE(device_history_list) v AS device_id
WHERE partition_date = '{yesterday_str}' WHERE partition_date = {yesterday_str}
)t1 )t1
JOIN JOIN
( (
SELECT device_id SELECT device_id
FROM online.ml_device_history_detail FROM online.ml_device_history_detail
WHERE partition_date = '{yesterday_str}' WHERE partition_date = {yesterday_str}
AND is_login_doctor = '1' AND is_login_doctor = '1'
)t2 )t2
ON t1.device_id = t2.device_id ON t1.device_id = t2.device_id
)t2 )t2
on t1.user_id=t2.user_id on t1.user_id=t2.user_id
group by partition_date,device_id""".format(yesterday_str=yesterday_str, today_str=today_str) group by partition_date,device_id
print(sql_dev_device_id) )dev
dev_df = spark.sql(sql_dev_device_id) on t1.partition_date=dev.partition_date and t1.cl_id=dev.device_id
dev_df_view = dev_df.createOrReplaceTempView("dev_view") WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='')
dev_df.cache() and (dev.device_id is null or dev.device_id ='')
dev_df.show(1) GROUP BY t1.partition_date,t2.active_type,device_os_type,channel
sql_res = dev_df.collect() )t
for res in sql_res: )t3
print(res)
print("-------------------------------")
sql_spam_pv_device_id = """
select device_id from (SELECT DISTINCT device_id
FROM ml.ml_d_ct_dv_devicespam_d --去除机构刷单设备，即作弊设备（浏览和曝光事件去除）
WHERE partition_day='{yesterday_str}'
UNION ALL
SELECT DISTINCT device_id
FROM dim.dim_device_user_staff)
--去除内网用户
""".format(yesterday_str=yesterday_str)
print(sql_spam_pv_device_id)
spam_pv_df = spark.sql(sql_spam_pv_device_id)
spam_pv_view = spam_pv_df.createOrReplaceTempView("spam_pv")
spam_pv_df.show(1)
sql_res = spam_pv_df.collect()
spam_pv_df.cache()
for res in sql_res:
print(res)
print("-------------------------------") LEFT JOIN
(--昨天搜索结果页卡片点击pv
SELECT t1.partition_date,active_type,device_os_type,channel
sql = r""" ,sum(hexin) as hexin_card_click_pv
SELECT t3.partition_date as partition_date ,sum(neirong) as neirong_card_click_pv
,t3.device_os_type as device_os_type FROM
,t3.active_type as active_type (
,t3.channel as channel SELECT NVL(t2.partition_date,t3.partition_date) as partition_date
,NVL(t3.search_pv,0) as pv ,NVL(t2.cl_id,t3.cl_id) as cl_id
,NVL(t3.search_uv,0) as uv ,NVL(t2.query,t3.query) as query
,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.hexin_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_core_pv ,NVL(t2.pv,0) as hexin
,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.neirong_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_pv ,NVL(t3.pv,0) as neirong
FROM FROM
(--昨天总搜索量 (--核心卡片点击
SELECT partition_date,active_type,device_os_type,channel,search_pv,search_uv SELECT partition_date
FROM ,params['query'] as query
( ,cl_id
SELECT t1.partition_date,active_type,device_os_type,channel ,count(1) as pv
,count(t1.cl_id) as search_pv FROM online.bl_hdfs_maidian_updates
,count(distinct t1.cl_id) as search_uv WHERE partition_date >= {yesterday_str}
FROM AND partition_date < {today_str}
( AND ((action in ('search_result_click_recommend_item','search_result_welfare_click_item','search_result_hospital_click_item','search_result_doctor_click_item','on_click_doctor_card', 'on_click_hospital_card')
SELECT partition_date AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor'))
,params['query'] as query or (action = 'goto_welfare_detail' AND params [ 'from' ] = 'search_result_welfare_recommend')
,cl_id or (action = 'on_click_card' AND params['card_content_type'] in ('service','hospital','doctor') AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor'))
FROM online.bl_hdfs_maidian_updates or (action = 'on_click_button' AND params['button_name'] = 'check_plan' AND page_name = 'search_result_more'))
WHERE partition_date >= {yesterday_str} GROUP BY partition_date
AND partition_date < {today_str} ,params['query']
AND action in ('do_search','search_result_click_search') ,cl_id
)t2
UNION ALL FULL JOIN
SELECT partition_date,params['query'] as query,cl_id (--内容卡片点击
FROM online.bl_hdfs_maidian_updates SELECT partition_date
WHERE partition_date >= {yesterday_str} ,params['query'] as query
AND partition_date < {today_str} ,cl_id
AND action = 'on_click_card' ,count(1) as pv
AND params['page_name']='search_home' FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
UNION ALL AND partition_date < {today_str}
SELECT partition_date AND ((action in ('on_click_topic_card','on_click_diary_card','search_result_click_infomation_item')
,params['card_name'] as query AND page_name in ('search_result_more','search_result_diary','search_result_post'))
,cl_id or (action = 'on_click_card' AND params['card_content_type'] in ('answer','diary') AND page_name in ('search_result_more','search_result_diary','search_result_question_answer')))
FROM online.bl_hdfs_maidian_updates GROUP BY partition_date
WHERE partition_date >= {yesterday_str} ,params['query']
AND partition_date < {today_str} ,cl_id
AND action = 'on_click_card' )t3
AND params['in_page_pos']='猜你喜欢' on t3.partition_date=t2.partition_date
AND params['tab_name']='精选' AND t3.query=t2.query
AND params['card_type']='search_word' AND t3.cl_id=t2.cl_id
--AND page_name='home' android的page_name为空 )t1
JOIN
UNION ALL (
SELECT partition_date SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND action = 'on_click_card'
AND page_name='welfare_home'
AND params['card_type'] ='search_word'
AND params['in_page_pos']='大家都在搜'
UNION ALL
SELECT partition_date
,params['card_name'] as query
,cl_id
FROM online.bl_hdfs_maidian_updates
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND int(split(app_version,'\\.')[1]) >= 27
AND action='on_click_card'
AND params['card_type']='highlight_word'
)t1
JOIN
(
SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type
FROM FROM
( (
SELECT SELECT
partition_date,m.device_id partition_date,m.device_id
,array(device_os_type ,'合计') as device_os_type ,array(device_os_type ,'合计') as device_os_type
,array(case WHEN active_type = '4' THEN '老活' ,array(case WHEN active_type = '4' THEN '老活'
WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type
,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel ,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel
FROM online.ml_device_day_active_status m FROM online.ml_device_day_active_status m
LEFT JOIN LEFT JOIN
(SELECT code,is_ai_channel,partition_day (SELECT code,is_ai_channel,partition_day
FROM DIM.DIM_AI_CHANNEL_ZP_NEW FROM DIM.DIM_AI_CHANNEL_ZP_NEW
WHERE partition_day>= {yesterday_str} WHERE partition_day>= {yesterday_str}
AND partition_day < {today_str}) tmp AND partition_day < {today_str}) tmp
ON m.partition_date=tmp.partition_day AND first_channel_source_type=code ON m.partition_date=tmp.partition_day AND first_channel_source_type=code
WHERE partition_date >= {yesterday_str} WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str} AND partition_date < {today_str}
AND active_type in ('1','2','4') AND active_type in ('1','2','4')
AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3' AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang' ,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1' ,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4' ,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100' ,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ' ,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
,'promotion_shike','promotion_julang_jl03','promotion_zuimei') ,'promotion_shike','promotion_julang_jl03','promotion_zuimei')
AND first_channel_source_type not like 'promotion\_jf\_%' AND first_channel_source_type not like 'promotion\_jf\_%'
) mas ) mas
LATERAL VIEW explode(mas.channel) t2 AS channel LATERAL VIEW explode(mas.channel) t2 AS channel
LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type
LATERAL VIEW explode(mas.active_type) t2 AS active_type LATERAL VIEW explode(mas.active_type) t2 AS active_type
)t2 )dev
on t1.cl_id=t2.device_id AND t1.partition_date = t2.partition_date on t1.cl_id=dev.device_id and t1.partition_date = dev.partition_date
LEFT JOIN
LEFT JOIN spam_pv on spam_pv.device_id=t1.cl_id (
LEFT JOIN dev_view SELECT DISTINCT device_id
on t1.partition_date=dev_view.partition_date and t1.cl_id=dev_view.device_id FROM ml.ml_d_ct_dv_devicespam_d --去除机构刷单设备，即作弊设备（浏览和曝光事件去除）
WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='') WHERE partition_day={yesterday_str}
and (dev_view.device_id is null or dev_view.device_id ='')
GROUP BY t1.partition_date,t2.active_type,device_os_type,channel UNION ALL
)t SELECT DISTINCT device_id
)t3 FROM dim.dim_device_user_staff --去除内网用户
)spam_pv
LEFT JOIN on spam_pv.device_id=t1.cl_id
(--昨天搜索结果页卡片点击pv LEFT JOIN
SELECT t1.partition_date,active_type,device_os_type,channel (
,sum(hexin) as hexin_card_click_pv SELECT partition_date,device_id
,sum(neirong) as neirong_card_click_pv FROM
FROM (--找出user_id当天活跃的第一个设备id
( SELECT user_id,partition_date,
SELECT NVL(t2.partition_date,t3.partition_date) as partition_date if(size(device_list) > 0, device_list [ 0 ], '') AS device_id
,NVL(t2.cl_id,t3.cl_id) as cl_id FROM online.ml_user_updates
,NVL(t2.query,t3.query) as query WHERE partition_date>='${{yesterday_str}}' AND partition_date<'${{today_str}}'
,NVL(t2.pv,0) as hexin )t1
,NVL(t3.pv,0) as neirong JOIN
FROM ( --医生账号
(--核心卡片点击 SELECT distinct user_id
SELECT partition_date FROM online.tl_hdfs_doctor_view
,params['query'] as query WHERE partition_date = {yesterday_str}
,cl_id
,count(1) as pv --马甲账号/模特用户
FROM online.bl_hdfs_maidian_updates UNION ALL
WHERE partition_date >= {yesterday_str} SELECT user_id
AND partition_date < {today_str} FROM ml.ml_c_ct_ui_user_dimen_d
AND ((action in ('search_result_click_recommend_item','search_result_welfare_click_item','search_result_hospital_click_item','search_result_doctor_click_item','on_click_doctor_card', 'on_click_hospital_card') WHERE partition_day = {yesterday_str}
AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor')) AND (is_puppet = 'true' or is_classifyuser = 'true')
or (action = 'goto_welfare_detail' AND params [ 'from' ] = 'search_result_welfare_recommend')
or (action = 'on_click_card' AND params['card_content_type'] in ('service','hospital','doctor') AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor')) UNION ALL
or (action = 'on_click_button' AND params['button_name'] = 'check_plan' AND page_name = 'search_result_more')) --公司内网覆盖用户
GROUP BY partition_date select distinct user_id
,params['query'] from dim.dim_device_user_staff
,cl_id
)t2 UNION ALL
FULL JOIN --登陆过医生设备
(--内容卡片点击 SELECT distinct t1.user_id
SELECT partition_date FROM
,params['query'] as query (
,cl_id SELECT user_id, v.device_id as device_id
,count(1) as pv FROM online.ml_user_history_detail
FROM online.bl_hdfs_maidian_updates LATERAL VIEW EXPLODE(device_history_list) v AS device_id
WHERE partition_date >= {yesterday_str} WHERE partition_date = {yesterday_str}
AND partition_date < {today_str} )t1
AND ((action in ('on_click_topic_card','on_click_diary_card','search_result_click_infomation_item') JOIN
AND page_name in ('search_result_more','search_result_diary','search_result_post')) (
or (action = 'on_click_card' AND params['card_content_type'] in ('answer','diary') AND page_name in ('search_result_more','search_result_diary','search_result_question_answer'))) SELECT device_id
GROUP BY partition_date FROM online.ml_device_history_detail
,params['query'] WHERE partition_date = {yesterday_str}
,cl_id AND is_login_doctor = '1'
)t3 )t2
on t3.partition_date=t2.partition_date ON t1.device_id = t2.device_id
AND t3.query=t2.query )t2
AND t3.cl_id=t2.cl_id on t1.user_id=t2.user_id
)t1 group by t1.partition_date,device_id
JOIN )dev1
( on t1.partition_date=dev1.partition_date and t1.cl_id=dev1.device_id
SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='')
FROM and (dev1.device_id is null or dev1.device_id ='')
( GROUP BY t1.partition_date,active_type,device_os_type,channel
SELECT )t4
partition_date,m.device_id on t3.partition_date=t4.partition_date and t3.active_type=t4.active_type and t3.device_os_type = t4.device_os_type AND t3.channel = t4.channel
,array(device_os_type ,'合计') as device_os_type
,array(case WHEN active_type = '4' THEN '老活'
WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type
,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel """.format(today_str=today_str, yesterday_str=yesterday_str, )
FROM online.ml_device_day_active_status m
LEFT JOIN
(SELECT code,is_ai_channel,partition_day
FROM DIM.DIM_AI_CHANNEL_ZP_NEW
WHERE partition_day>= {yesterday_str}
AND partition_day < {today_str}) tmp
ON m.partition_date=tmp.partition_day AND first_channel_source_type=code
WHERE partition_date >= {yesterday_str}
AND partition_date < {today_str}
AND active_type in ('1','2','4')
AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
,'promotion_shike','promotion_julang_jl03','promotion_zuimei')
AND first_channel_source_type not like 'promotion\_jf\_%'
) mas
LATERAL VIEW explode(mas.channel) t2 AS channel
LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type
LATERAL VIEW explode(mas.active_type) t2 AS active_type
)dev
on t1.cl_id=dev.device_id and t1.partition_date = dev.partition_date
LEFT JOIN spam_pv on spam_pv.device_id=t1.cl_id
LEFT JOIN dev_view
on t1.partition_date=dev_view.partition_date and t1.cl_id=dev_view.device_id
WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='')
and (dev_view.device_id is NULL or dev_view.device_id ='')
GROUP BY t1.partition_date,active_type,device_os_type,channel
)t4
on t3.partition_date=t4.partition_date and t3.active_type=t4.active_type and t3.device_os_type = t4.device_os_type AND t3.channel = t4.channel
""".format(today_str=today_str, yesterday_str=yesterday_str)
device_df = spark.sql(sql) device_df = spark.sql(sql)
device_df.show(1, False) device_df.show(1, False)
sql_res = device_df.collect() sql_res = device_df.collect()
...@@ -379,372 +719,34 @@ on t3.partition_date=t4.partition_date and t3.active_type=t4.active_type and t3. ...@@ -379,372 +719,34 @@ on t3.partition_date=t4.partition_date and t3.active_type=t4.active_type and t3.
print(res) print(res)
device_df.createOrReplaceTempView("data_table") device_df.createOrReplaceTempView("data_table")
# sql = r""" collects_sql = """
# SELECT t3.partition_date as partition_date SELECT device_type,active_type,channel_type,ROUND(if(NVL(sum(uv),0) <> 0 ,NVL(sum(search_core_pv),0)/NVL(sum(uv),0) ,0),5) as core_pv_division_uv,
# ,t3.device_os_type as device_os_type ROUND(if(NVL(sum(uv),0) <> 0 ,NVL(sum(search_pv),0)/NVL(sum(uv),0) , 0),5) as pv_division_uv
# ,t3.active_type as active_type FROM data_table GROUP BY device_type,active_type,channel_type
# ,t3.channel as channel """
# ,NVL(t3.search_pv,0) as pv finnal_df = spark.sql(collects_sql)
# ,NVL(t3.search_uv,0) as uv
# ,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.hexin_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_core_pv finnal_df.show(1, False)
# ,if(NVL(t3.search_uv,0) <> 0 ,concat(cast((NVL(t4.neirong_card_click_pv,0)/NVL(t3.search_uv,0)) as decimal(18,2)),'') , '-') as search_pv sql_res = finnal_df.collect()
# FROM for res in sql_res:
# (--昨天总搜索量 # print(res)
# SELECT partition_date,active_type,device_os_type,channel,search_pv,search_uv device_type = res.device_type
# FROM active_type = res.active_type
# ( channel_type = res.channel_type
# SELECT t1.partition_date,active_type,device_os_type,channel core_pv_division_uv = res.core_pv_division_uv
# ,count(t1.cl_id) as search_pv pv_division_uv = res.pv_division_uv
# ,count(distinct t1.cl_id) as search_uv pid = hashlib.md5(
# FROM (today_str + device_type + active_type + channel_type).encode("utf8")).hexdigest()
# ( instert_sql = """replace into search_strategy_d(
# SELECT partition_date day_id,device_type,active_type,channel_type,core_pv_division_uv,pv_division_uv,pid
# ,params['query'] as query ) VALUES('{day_id}','{device_type}','{active_type}','{channel_type}',{core_pv_division_uv},{pv_division_uv},'{pid}');""".format(
# ,cl_id day_id=today_str, device_type=device_type,
# FROM online.bl_hdfs_maidian_updates active_type=active_type, channel_type=channel_type, core_pv_division_uv=core_pv_division_uv,pv_division_uv=pv_division_uv,pid=pid
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str} )
# AND action in ('do_search','search_result_click_search') print(instert_sql)
# # cursor.execute("set names 'UTF8'")
# UNION ALL res = cursor.execute(instert_sql)
# SELECT partition_date,params['query'] as query,cl_id db.commit()
# FROM online.bl_hdfs_maidian_updates print(res)
# WHERE partition_date >= {yesterday_str} db.close()
# AND partition_date < {today_str}
# AND action = 'on_click_card'
# AND params['page_name']='search_home'
#
# UNION ALL
# SELECT partition_date
# ,params['card_name'] as query
# ,cl_id
# FROM online.bl_hdfs_maidian_updates
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str}
# AND action = 'on_click_card'
# AND params['in_page_pos']='猜你喜欢'
# AND params['tab_name']='精选'
# AND params['card_type']='search_word'
# --AND page_name='home' android的page_name为空
#
# UNION ALL
# SELECT partition_date
# ,params['card_name'] as query
# ,cl_id
# FROM online.bl_hdfs_maidian_updates
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str}
# AND action = 'on_click_card'
# AND page_name='welfare_home'
# AND params['card_type'] ='search_word'
# AND params['in_page_pos']='大家都在搜'
#
# UNION ALL
# SELECT partition_date
# ,params['card_name'] as query
# ,cl_id
# FROM online.bl_hdfs_maidian_updates
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str}
# AND int(split(app_version,'\\.')[1]) >= 27
# AND action='on_click_card'
# AND params['card_type']='highlight_word'
# )t1
# JOIN
# (
# SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type
# FROM
# (
# SELECT
# partition_date,m.device_id
# ,array(device_os_type ,'合计') as device_os_type
# ,array(case WHEN active_type = '4' THEN '老活'
# WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type
# ,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel
# FROM online.ml_device_day_active_status m
# LEFT JOIN
# (SELECT code,is_ai_channel,partition_day
# FROM DIM.DIM_AI_CHANNEL_ZP_NEW
# WHERE partition_day>= {yesterday_str}
# AND partition_day < {today_str}) tmp
# ON m.partition_date=tmp.partition_day AND first_channel_source_type=code
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str}
# AND active_type in ('1','2','4')
# AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
# ,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
# ,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
# ,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
# ,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
# ,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
# ,'promotion_shike','promotion_julang_jl03','promotion_zuimei')
# AND first_channel_source_type not like 'promotion\_jf\_%'
# ) mas
# LATERAL VIEW explode(mas.channel) t2 AS channel
# LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type
# LATERAL VIEW explode(mas.active_type) t2 AS active_type
# )t2
# on t1.cl_id=t2.device_id AND t1.partition_date = t2.partition_date
#
# LEFT JOIN
# (
# SELECT DISTINCT device_id
# FROM ml.ml_d_ct_dv_devicespam_d --去除机构刷单设备，即作弊设备（浏览和曝光事件去除）
# WHERE partition_day={yesterday_str}
#
# UNION ALL
# SELECT DISTINCT device_id
# FROM dim.dim_device_user_staff --去除内网用户
# )spam_pv
# on spam_pv.device_id=t1.cl_id
# LEFT JOIN
# (
# SELECT partition_date,device_id
# FROM
# (--找出user_id当天活跃的第一个设备id
# SELECT user_id,partition_date,
# if(size(device_list) > 0, device_list [ 0 ], '') AS device_id
# FROM online.ml_user_updates
# WHERE partition_date>='${{yesterday_str}}' AND partition_date<'${{today_str}}'
# )t1
# JOIN
# ( --医生账号
# SELECT distinct user_id
# FROM online.tl_hdfs_doctor_view
# WHERE partition_date = {yesterday_str}
#
# --马甲账号/模特用户
# UNION ALL
# SELECT user_id
# FROM ml.ml_c_ct_ui_user_dimen_d
# WHERE partition_day = {yesterday_str}
# AND (is_puppet = 'true' or is_classifyuser = 'true')
#
# UNION ALL
# --公司内网覆盖用户
# select distinct user_id
# from dim.dim_device_user_staff
#
# UNION ALL
# --登陆过医生设备
# SELECT distinct t1.user_id
# FROM
# (
# SELECT user_id, v.device_id as device_id
# FROM online.ml_user_history_detail
# LATERAL VIEW EXPLODE(device_history_list) v AS device_id
# WHERE partition_date = {yesterday_str}
# )t1
# JOIN
# (
# SELECT device_id
# FROM online.ml_device_history_detail
# WHERE partition_date = {yesterday_str}
# AND is_login_doctor = '1'
# )t2
# ON t1.device_id = t2.device_id
# )t2
# on t1.user_id=t2.user_id
# group by partition_date,device_id
# )dev
# on t1.partition_date=dev.partition_date and t1.cl_id=dev.device_id
# WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='')
# and (dev.device_id is null or dev.device_id ='')
# GROUP BY t1.partition_date,t2.active_type,device_os_type,channel
# )t
# )t3
#
# LEFT JOIN
# (--昨天搜索结果页卡片点击pv
# SELECT t1.partition_date,active_type,device_os_type,channel
# ,sum(hexin) as hexin_card_click_pv
# ,sum(neirong) as neirong_card_click_pv
# FROM
# (
# SELECT NVL(t2.partition_date,t3.partition_date) as partition_date
# ,NVL(t2.cl_id,t3.cl_id) as cl_id
# ,NVL(t2.query,t3.query) as query
# ,NVL(t2.pv,0) as hexin
# ,NVL(t3.pv,0) as neirong
# FROM
# (--核心卡片点击
# SELECT partition_date
# ,params['query'] as query
# ,cl_id
# ,count(1) as pv
# FROM online.bl_hdfs_maidian_updates
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str}
# AND ((action in ('search_result_click_recommend_item','search_result_welfare_click_item','search_result_hospital_click_item','search_result_doctor_click_item','on_click_doctor_card', 'on_click_hospital_card')
# AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor'))
# or (action = 'goto_welfare_detail' AND params [ 'from' ] = 'search_result_welfare_recommend')
# or (action = 'on_click_card' AND params['card_content_type'] in ('service','hospital','doctor') AND page_name in ('search_result_more','search_result_welfare','search_result_hospital','search_result_doctor'))
# or (action = 'on_click_button' AND params['button_name'] = 'check_plan' AND page_name = 'search_result_more'))
# GROUP BY partition_date
# ,params['query']
# ,cl_id
# )t2
# FULL JOIN
# (--内容卡片点击
# SELECT partition_date
# ,params['query'] as query
# ,cl_id
# ,count(1) as pv
# FROM online.bl_hdfs_maidian_updates
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str}
# AND ((action in ('on_click_topic_card','on_click_diary_card','search_result_click_infomation_item')
# AND page_name in ('search_result_more','search_result_diary','search_result_post'))
# or (action = 'on_click_card' AND params['card_content_type'] in ('answer','diary') AND page_name in ('search_result_more','search_result_diary','search_result_question_answer')))
# GROUP BY partition_date
# ,params['query']
# ,cl_id
# )t3
# on t3.partition_date=t2.partition_date
# AND t3.query=t2.query
# AND t3.cl_id=t2.cl_id
# )t1
# JOIN
# (
# SELECT partition_date,device_id,t2.active_type,t2.channel,t2.device_os_type
# FROM
# (
# SELECT
# partition_date,m.device_id
# ,array(device_os_type ,'合计') as device_os_type
# ,array(case WHEN active_type = '4' THEN '老活'
# WHEN active_type in ('1','2') then '新增' END ,'合计') as active_type
# ,array(CASE WHEN is_ai_channel = 'true' THEN 'AI' ELSE '其他' END , '合计') as channel
# FROM online.ml_device_day_active_status m
# LEFT JOIN
# (SELECT code,is_ai_channel,partition_day
# FROM DIM.DIM_AI_CHANNEL_ZP_NEW
# WHERE partition_day>= {yesterday_str}
# AND partition_day < {today_str}) tmp
# ON m.partition_date=tmp.partition_day AND first_channel_source_type=code
# WHERE partition_date >= {yesterday_str}
# AND partition_date < {today_str}
# AND active_type in ('1','2','4')
# AND first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
# ,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
# ,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
# ,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
# ,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
# ,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
# ,'promotion_shike','promotion_julang_jl03','promotion_zuimei')
# AND first_channel_source_type not like 'promotion\_jf\_%'
# ) mas
# LATERAL VIEW explode(mas.channel) t2 AS channel
# LATERAL VIEW explode(mas.device_os_type) t2 AS device_os_type
# LATERAL VIEW explode(mas.active_type) t2 AS active_type
# )dev
# on t1.cl_id=dev.device_id and t1.partition_date = dev.partition_date
# LEFT JOIN
# (
# SELECT DISTINCT device_id
# FROM ml.ml_d_ct_dv_devicespam_d --去除机构刷单设备，即作弊设备（浏览和曝光事件去除）
# WHERE partition_day={yesterday_str}
#
# UNION ALL
# SELECT DISTINCT device_id
# FROM dim.dim_device_user_staff --去除内网用户
# )spam_pv
# on spam_pv.device_id=t1.cl_id
# LEFT JOIN
# (
# SELECT partition_date,device_id
# FROM
# (--找出user_id当天活跃的第一个设备id
# SELECT user_id,partition_date,
# if(size(device_list) > 0, device_list [ 0 ], '') AS device_id
# FROM online.ml_user_updates
# WHERE partition_date>='${{yesterday_str}}' AND partition_date<'${{today_str}}'
# )t1
# JOIN
# ( --医生账号
# SELECT distinct user_id
# FROM online.tl_hdfs_doctor_view
# WHERE partition_date = {yesterday_str}
#
# --马甲账号/模特用户
# UNION ALL
# SELECT user_id
# FROM ml.ml_c_ct_ui_user_dimen_d
# WHERE partition_day = {yesterday_str}
# AND (is_puppet = 'true' or is_classifyuser = 'true')
#
# UNION ALL
# --公司内网覆盖用户
# select distinct user_id
# from dim.dim_device_user_staff
#
# UNION ALL
# --登陆过医生设备
# SELECT distinct t1.user_id
# FROM
# (
# SELECT user_id, v.device_id as device_id
# FROM online.ml_user_history_detail
# LATERAL VIEW EXPLODE(device_history_list) v AS device_id
# WHERE partition_date = {yesterday_str}
# )t1
# JOIN
# (
# SELECT device_id
# FROM online.ml_device_history_detail
# WHERE partition_date = {yesterday_str}
# AND is_login_doctor = '1'
# )t2
# ON t1.device_id = t2.device_id
# )t2
# on t1.user_id=t2.user_id
# group by t1.partition_date,device_id
# )dev1
# on t1.partition_date=dev1.partition_date and t1.cl_id=dev1.device_id
# WHERE (spam_pv.device_id IS NULL or spam_pv.device_id ='')
# and (dev1.device_id is null or dev1.device_id ='')
# GROUP BY t1.partition_date,active_type,device_os_type,channel
# )t4
# on t3.partition_date=t4.partition_date and t3.active_type=t4.active_type and t3.device_os_type = t4.device_os_type AND t3.channel = t4.channel
#
#
#
# """.format(today_str=today_str, yesterday_str=yesterday_str, )
# device_df = spark.sql(sql)
# device_df.show(1, False)
# sql_res = device_df.collect()
# for res in sql_res:
# print(res)
# device_df.createOrReplaceTempView("data_table")
# collects_sql = """
# SELECT device_type,active_type,channel_type,ROUND(if(NVL(sum(uv),0) <> 0 ,NVL(sum(search_core_pv),0)/NVL(sum(uv),0) ,0),5) as core_pv_division_uv,
# ROUND(if(NVL(sum(uv),0) <> 0 ,NVL(sum(search_pv),0)/NVL(sum(uv),0) , 0),5) as pv_division_uv
# FROM data_table GROUP BY device_type,active_type,channel_type
# """
# finnal_df = spark.sql(collects_sql)
# finnal_df.show(1, False)
# sql_res = finnal_df.collect()
# for res in sql_res:
# # print(res)
# device_type = res.device_type
# active_type = res.active_type
# channel_type = res.channel_type
# core_pv_division_uv = res.core_pv_division_uv
# pv_division_uv = res.pv_division_uv
# pid = hashlib.md5(
# (today_str + device_type + active_type + channel_type).encode("utf8")).hexdigest()
# instert_sql = """replace into search_strategy_d(
# day_id,device_type,active_type,channel_type,core_pv_division_uv,pv_division_uv,pid
# ) VALUES('{day_id}','{device_type}','{active_type}','{channel_type}',{core_pv_division_uv},{pv_division_uv},'{pid}');""".format(
# day_id=today_str, device_type=device_type,
# active_type=active_type, channel_type=channel_type, core_pv_division_uv=core_pv_division_uv,pv_division_uv=pv_division_uv,pid=pid
#
# )
# print(instert_sql)
# # cursor.execute("set names 'UTF8'")
# res = cursor.execute(instert_sql)
# db.commit()
# print(res)
# db.close()
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment