As I mentioned in my previous post, I have been playing with Tableau Desktop in my spare time. I am just experimenting with it and there are no structured instructions that I am following. In this post, I intend to connect to Excel data source and do some analysis. I chose the sample excel provided by Tableau which is mostly located in “My Documents\My Tableau Repository\Datasources\Sample – Superstore Sales (Excel).xls”. Once connected, Tableau will try to create Dimensions and Measures for you. It seems like, all the numeric field are converted to measures and string fields are converted to dimensions. To test my theory, I added two columns in the source excel. One called Sales Category with string values and other Weight with numeric values. Sure enough Sales Category became a dimension and Weight became a measure.
Strings are converted to dimensions and numeric values are converted to measures
You will also observe a small icon beside each field. The icon kind of represents the data type. For e.g. ‘Abc’ means it’s a character data, a # represents numeric field. There are certain special icons. For e.g. a little globe means the field has something to do with geography data, small square with clock is date/time field, that chart like thing is a bin(more on this later) and pin is Group (again more on that later). You can change data type by right clicking on the field and selecting Change Data Type. You can move stuff from measures to dimensions and vice a versa (of course if it makes sense!). For e.g. Order Id in measures does not make much sense, you can drag it to dimension panel and it will become a dimension. My gut feeling is that measure is treated just like another dimension, something similar to what SQL Server 2012 Tabular does. Creating a hierarchy is simple enough. Right click on a field from dimension panel and select create hierarchy. Drag the fields under the hierarchy name.
Coming to the bins and groups part mentioned above. I was a bit puzzled and struggled to get my head around the concept. Once I got hang of it, its is one of my favourite features. Bins are like buckets of data or data groups. For e.g you have a persons age as measure, you can group it in bins of 0-10,10-20,20-30 and so on or as ‘Kids’,’Adults’ etc. This tutorial is a good starting point.
That’s it for this post. In the next post, I am going to try to answer some questions based on some interesting datasets I have found using Tableau. So stay tuned!!