Most data sets with this appearance are structured so each row is an independent observation, which could lead to misinterpretation if that’s not the case.Ģ. ĭata structured like the table above is also potentially misleading because it would be easy to assume that, as there were 18 grey blankets in 2015, there were 52,056 ants seen on those 18 grey blankets. To get around this, we’d need to use a different aggregation (such as average or min), or use an LOD expression to fix the number of ants per year, thus preventing accidental over-counting. This means that, in analysis, if we simply brought out Year and Number of Ants (as a sum) we’d quadruple the number of ants. Structuring the data this way isn’t perfect because we have the number of ants repeated for each year. And the resulting data would look like this: Repeat the number of ants seen for every blanket color: If we were to join these tables, we could simply replicate the number of ants seen in a given year for each row in the blanket table. If we don't like either of the options above, we can choose to not join the tables and perform the analyses with the two tables independently.ġ.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |