如何在SSAS中保留数据点的历史值
我们目前有一份报告，显示每周数据点值的变化。它存储在数据仓库中，其中有一个“快照”字段，显示这些值的日期拉取。该日期还可以映射到“快照”维度，以获取其他快照相关信息（拉动类型的标志，是当前拉动还是上一次拉动等）
关于多维数据集开发（ssas），这是一个常见的情况实践还是有处理这种情况的机制？
we currently have a report that displays a data point value change week to week. This is stored in a data warehouse with a 'snapshot' field showing the date pull of those values. This date can also be mapped to a 'snapshot' dimension for other snapshot related information (flags on what kind of pull it was, whether it's the current pull or the previous pull etcetera)
In regard to cube development (ssas) is this a common practice or is there a mechanism for handling this scenario?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
通常，快照数据存储在其自己的定期快照事实表中（请参阅 Kimball 方法）。然后，在 SSAS 数据库中，事实表作为度量值组引入（就像其他事实表一样）。唯一的区别是为度量指定的聚合函数...通常快照事实表中的事实不是完全相加的，因此您不能使用 SUM() 作为聚合函数。最简单的思考方法是银行账户余额……您无法将所有月份的余额相加以获得年底的余额。
Typically the snapshot data is stored in its own periodic snapshot fact table (see Kimball methodology). Then in the SSAS database, the fact table is brought in as a measure group (just like the other fact tables). The only difference is going to be the aggregation function specified for the measures...typically facts in a snapshot fact table are not fully additive so you can't use SUM() as the aggregation function. The easiest way to think about it is in terms of a bank account balance...you can't add the balance across all the months to get the balance at the end of the year.