Here's a quick thought I'd like to share:
A big part of my job as a consultant is supporting the development of other data analysts, especially as it relates to Tableau and Alteryx. I have found that when it comes to data preparation many people, including myself, can struggle with joins, pivoting, and aggregating data. Not only can it be difficult to conceptualize the changes that happen to our data model, but these actions can change the granularity of the row level.
Understanding what the row-level of the data is is critical to understanding how your data is structured, and what you need to do to prepare it. Say we are given some financial performance data about Walmarts across the United States. The immediate questions you should be asking are:
- Is there one row per store?
- What is the granularity of the time? Is there a row for every month or quarter? Or maybe this is annual data?
- If there are multiple rows of the same store, how does this impact aggregations when we visualize the data?
When I am assisting someone in preparing their data, I will always define what I understand to be the row level before we get into fixing the issues. Say it aloud to your stakeholder and have them confirm you are correct, but also validate the row level yourself through data discovery. The data can be hundreds of columns and confusing as hell! If you understand your row level everything gets easier.
