Affinity Water Domestic Water Quality 2022

Data Origin Samples were taken from customer taps. They were then analysed, and the results were uploaded to a sampling system. This dataset is an extract from this database. Data Triage We discussed: Whether to use individual samples or averages What date range would be appropriate for MVP How best to anonymise the data geographically Preferred: Geospatial to LSOA Do geospatial join to find the LSOA that each customer sample location (within whatever coordinate system you have them as) is found within Acceptable for MVP: Postcode to LSOA Join ONS dataset on postcode, keep LSOA code field Why is Geospatial to LSOA preferred? Because it is more accurate since the postcode mapping is best fitted by plotting the location of the postcode's mean address rather than sample point’s specific location.### 数据来源 样本取自客户采样点。随后对样本开展分析，并将分析结果上传至采样系统。本数据集系从该数据库中抽取的子集。 ### 数据梳理 我们曾就以下议题展开研讨： 1. 应采用单一样本还是样本均值 2. 适配最小可行产品(MVP)的合适日期范围为何 3. 如何在地理维度实现最优的数据匿名化 #### 首选方案 将地理空间点位匹配至低层级输出区域(LSOA)：执行地理空间关联操作，以确定每个客户样本点位（无论其采用何种坐标系统）所属的LSOA区域。 #### MVP适配可接受方案 通过邮编与LSOA区域进行关联：调取英国国家统计局(ONS)的邮编数据集，保留其中的LSOA编码字段。 #### 方案优选原因 为何优先选择地理空间点位匹配至LSOA的方案？原因在于该方案精度更高：邮编映射是通过拟合邮编对应地址的平均点位完成，而非直接匹配样本点位的精确坐标。