Merge pull request #20 from aws-samples/intent

change some code

README.md

@@ -67,7 +67,7 @@ sudo su - ec2-user
 
 ```bash  
 # Install components
-sudo yum install docker python3-pip git -y && pip3 install -U awscli && pip3 install docker-compose
+sudo dnf install docker python3-pip git -y && pip3 install -U awscli && pip3 install docker-compose
 
 # For Amazon Linux 2，use yum to replace dnf
 

README_CN.md

@@ -64,7 +64,7 @@ sudo su - ec2-user
 
 ```bash
 # 安装组件
-sudo yum install docker python3-pip git -y && pip3 install -U awscli && pip3 install docker-compose
+sudo dnf install docker python3-pip git -y && pip3 install -U awscli && pip3 install docker-compose
 
 # 对于 Amazon Linux 2，可以使用yum 替换 dnf
 

application/Index.py

@@ -5,7 +5,7 @@
     page_icon="👋",
 )
 
-st.write("## Welcome to Intelligent BI Playground! Aoyu👋")
+st.write("## Welcome to Intelligent BI Playground!👋")
 
 st.sidebar.success("Select a demo above.")
 

application/utils/prompts/guidance_prompt.py

@@ -4,26 +4,7 @@
 """
 
 guidance_prompt_dict['sonnet-20240229v1-0'] = """
-divide actual value by target value to get A/T% or 达成率.
-divide this year's sales by last year's sales and then subtract one to get YTD YoY GR% or YTD 销量增长.
-divide this year's market sales by last year's market sales and then subtract one to get YoY MKT GR% or 市场增长.
-divide product sales by total sales to get Con% or 销售贡献.
-divide selected product sales by default market sales to get YTD YoY MS% or YTD市场占比.
-subtract last year's MS% from this year's MS% to get YTD YoY Δ MS% or YTD 市场占比变化.
-subtract last year's sales from this year's sales to sales incremental or 销量增量.
-sum last year's actual sales to get LY Actual or 去年实际销量.
-rank sales within selected products and selected channels to get Sale Value Rank or 销量排名.
-divide all increments by product increments to get YTD GR Con% or 销量贡献增长.
-sum MTD DDI Actual to get MTD Actual or 当月累计销量.
-divide the total amount of MTDD DDI Actual by the total amount of Target to get DDI A/T or DDI累计达成.
-calculate R3M GR% or 滚动3个月增长 step by step using the following four steps:
-step one: collect at least two years data earlier than the mentioned time in query.
-step two: calculate the sum of sales volume from the max month of existing data to the previous 3 months in the year mentioned in query as pre_r3m.
-step three: calculate the sum of sales volume relative to the same months of the previous year in step two as whole_r3m.
-step four: divide pre_r3m by whole_r3m and minus by one.
-use MYSQL Sytax, for example:
-use EXTRACT instead of DATE_TRUNC
-use EXTRACT instead of DATE_PART
+
 """
 
 class GuidancePromptMapper:

application/utils/prompts/table_prompt.py

@@ -1,142 +1,10 @@
 table_prompt_dict = {}
 table_prompt_dict['haiku-20240307v1-0']="""
-Event：存储的是一家药企的举办市场活动/会议的信息
-- name: trty_name, datatype: VARCHAR, comment: 岗位名称
-  annotataion:
-- name: event_name, datatype: VARCHAR, comment: 活动名称
-  annotataion:
-- name: event_date, datatype: DATE, comment: 活动时间
-  annotataion:
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion: 示例数据，Forxiga 中文名称是安达唐, Tagrisso,Breztri Aero
-- name: event_city, datatype: VARCHAR, comment: 活动城市
-  annotataion:
-- name: event_type, datatype: VARCHAR, comment: 活动类型
-  annotataion:
-- name: event_status, datatype: VARCHAR, comment: 活动状态
-  annotataion: 示例数据，Event Approved for Closing,Event Blocked,Event Cancelled,Event Rejected,Under Rework for Closing,Available for Closing,New,Under Rework,Event Submitted for Closing,Event Submitted
- 
-internal_sales: 是内部销量表(药品在内部代理和销售大区的销售数据), 存储销售岗位名称、产品、部门、买方所在省、买方所在区域、销售金额等信息, 如果涉及销售区域和买家可以在这里查询。
-- name: sales_date, datatype: DATE, comment: 销售日期
-  annotataion:
-- name: trty_name, datatype: VARCHAR, comment: 销售岗位名称
-  annotataion:
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion: 示例数据为,Forxiga(中文名称安达唐)，Tagrisso(中文名称泰瑞沙)，Breztri Aero(中文名称倍择瑞)
-- name: bu, datatype: VARCHAR, comment: 销售部门
-  annotataion: 示例数据为,RI&B, RGI,OBU,EC,Eagle,CVRM,CHC,BBU_County
-- name: sub_bu, datatype: VARCHAR, comment: 销售子部门
-  annotataion: 示例数据为,LC,CHC,BBU_County,IPT,GNR,OBU_County,EC
-- name: province_name, datatype: VARCHAR, comment: 买方所在省, 值是中国所有省份,如广东
-  annotataion:
-- name: area_name, datatype: VARCHAR, comment: 买方所在区
-  annotataion: 示例数据为,东部地区,中部地区,北部地区,南部地区,西部地区
-- name: region_name, datatype: VARCHAR, comment: 销售大区
-  annotataion:
-- name: region_center_name, datatype: VARCHAR, comment: 销售大区中心
-  annotataion:
-- name: inst_name, datatype: VARCHAR, comment: 买方名称,值大多是各类购买药品的医院/诊所名称
-  annotataion:
-- name: sales_val, datatype: DECIMAL, comment: 销售金额
-  annotataion:
-- name: target_val, datatype: DECIMAL, comment: 目标销售金额
-  annotataion:
- 
-external_sales: 是外部销量表(药品在市场上销售的数据), 存储销量来源、市场、产品、销售日期、销售金额等字段, 如果涉及市场排名可以在这里查询。存储的是多家药企的外部销售数据, 一个药企可能销售多种药, 每种药每段时间在市场上销量为一行数据。
- 
-- name: source, datatype: VARCHAR, comment: 数据来源, 只有'aia','ims'两种值
-  annotataion:
-- name: market, datatype: VARCHAR, comment: 市场
-  annotataion:
-- name: manufacturer, datatype: VARCHAR, comment:
-  annotataion:
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion:
-- name: is_az, datatype: VARCHAR, comment: 是否是AZ的产品
-  annotataion: 数值只有:Y或者N
-- name: sales_date, datatype: DATE, comment: 销售日期
-  annotataion: 日期的格式是2023-01-01
-- name: sales_val, datatype: DECIMAL, comment: 销售金额, 数据类型是(numeric)
-  annotataion:
- 
-listed_brand: 存储的是药企的客户每次来采购药品的数据
- 
-- name: inst_name, datatype: VARCHAR, comment: 买方名称
-  annotataion:
-- name: list_date, datatype: DATE, comment: 采购时间
-  annotataion: 示例时间是，2023-02-01
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion: 示例数据包括Forxiga,Tagrisso,Breztri Aerok
+
 """
 
 table_prompt_dict['sonnet-20240229v1-0']="""
-Event：存储的是一家药企的举办市场活动/会议的信息
-- name: trty_name, datatype: VARCHAR, comment: 岗位名称
-  annotataion:
-- name: event_name, datatype: VARCHAR, comment: 活动名称
-  annotataion:
-- name: event_date, datatype: DATE, comment: 活动时间
-  annotataion:
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion: 示例数据，Forxiga 中文名称是安达唐, Tagrisso,Breztri Aero
-- name: event_city, datatype: VARCHAR, comment: 活动城市
-  annotataion:
-- name: event_type, datatype: VARCHAR, comment: 活动类型
-  annotataion:
-- name: event_status, datatype: VARCHAR, comment: 活动状态
-  annotataion: 示例数据，Event Approved for Closing,Event Blocked,Event Cancelled,Event Rejected,Under Rework for Closing,Available for Closing,New,Under Rework,Event Submitted for Closing,Event Submitted
- 
-internal_sales: 是内部销量表(药品在内部代理和销售大区的销售数据), 存储销售岗位名称、产品、部门、买方所在省、买方所在区域、销售金额等信息, 如果涉及销售区域和买家可以在这里查询。
-- name: sales_date, datatype: DATE, comment: 销售日期
-  annotataion:
-- name: trty_name, datatype: VARCHAR, comment: 销售岗位名称
-  annotataion:
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion: 示例数据为,Forxiga(中文名称安达唐)，Tagrisso(中文名称泰瑞沙)，Breztri Aero(中文名称倍择瑞)
-- name: bu, datatype: VARCHAR, comment: 销售部门
-  annotataion: 示例数据为,RI&B, RGI,OBU,EC,Eagle,CVRM,CHC,BBU_County
-- name: sub_bu, datatype: VARCHAR, comment: 销售子部门
-  annotataion: 示例数据为,LC,CHC,BBU_County,IPT,GNR,OBU_County,EC
-- name: province_name, datatype: VARCHAR, comment: 买方所在省, 值是中国所有省份,如广东
-  annotataion:
-- name: area_name, datatype: VARCHAR, comment: 买方所在区
-  annotataion: 示例数据为,东部地区,中部地区,北部地区,南部地区,西部地区
-- name: region_name, datatype: VARCHAR, comment: 销售大区
-  annotataion:
-- name: region_center_name, datatype: VARCHAR, comment: 销售大区中心
-  annotataion:
-- name: inst_name, datatype: VARCHAR, comment: 买方名称,值大多是各类购买药品的医院/诊所名称
-  annotataion:
-- name: sales_val, datatype: DECIMAL, comment: 销售金额
-  annotataion:
-- name: target_val, datatype: DECIMAL, comment: 目标销售金额
-  annotataion:
- 
-external_sales: 是外部销量表(药品在市场上销售的数据), 存储销量来源、市场、产品、销售日期、销售金额等字段, 如果涉及市场排名可以在这里查询。存储的是多家药企的外部销售数据, 一个药企可能销售多种药, 每种药每段时间在市场上销量为一行数据。
- 
-- name: source, datatype: VARCHAR, comment: 数据来源, 只有'aia','ims'两种值
-  annotataion:
-- name: market, datatype: VARCHAR, comment: 市场
-  annotataion:
-- name: manufacturer, datatype: VARCHAR, comment:
-  annotataion:
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion:
-- name: is_az, datatype: VARCHAR, comment: 是否是AZ的产品
-  annotataion: 数值只有:Y或者N
-- name: sales_date, datatype: DATE, comment: 销售日期
-  annotataion: 日期的格式是2023-01-01
-- name: sales_val, datatype: DECIMAL, comment: 销售金额, 数据类型是(numeric)
-  annotataion:
- 
-listed_brand: 存储的是药企的客户每次来采购药品的数据
- 
-- name: inst_name, datatype: VARCHAR, comment: 买方名称
-  annotataion:
-- name: list_date, datatype: DATE, comment: 采购时间
-  annotataion: 示例时间是，2023-02-01
-- name: brand, datatype: VARCHAR, comment: 产品
-  annotataion: 示例数据包括Forxiga,Tagrisso,Breztri Aerok
+
 """
 
 class TablePromptMapper: