DS 39 had another training session with Carl Allchin in which, he stressed the importance of knowing our terminology to improve our understanding on how to use the software. So, in this blog post - I am going to recap everything we learnt so far today.
Saving different file types:
- Workbooks (.twb) - Tableau workbook files, can hold multiple worksheets and dashboards, without the dataset, no data points will be seen
- Packaged Workbooks (.twbx) - a single zip file of the workbook as well as the dataset, beneficial for sharing and working offline
- Data Source (.tds) - shortcut file to connect to the necessary data which connects to the actual dataset with any modifications made
- Packaged Data Source (.tdsx) - a zip file that contain the data source file, sharing this file mean other users do not need access to the original dataset
- Extract (.hyper) - a local file with the entire/partial dataset saved to your own area which, you can work with offline, this format is better for optimization and performance
Discrete vs Continuous:
When comparing measures and categories from Tableau Prep to dimensions and measures in Tableau Desktop, we can generally say:
Discrete = Categories/Dimensions and Continuous = Measures
However, this is not necessarily the case as we can have discrete dimensions and discrete measures or continuous dimensions and continuous measures.
Live vs Extract:
An extract of the dataset is essentially a snapshot of the dataset. With live datasets (let's use weather data as an example), they update constantly which, can change your workbooks. Taking an extract prevents this change to your workbooks. Extract files should not be sent to people who do not have permission to access the dataset otherwise... you may end up in jail.
Filtering:
- General filtering - gives a check box option list to which, you can include or exclude options
- Wildcard - filter through values depending on the letters with the options of 'starts with', 'ends with' and 'exactly matches'
- Condition - filter with a formula which can be computed through Tableau such as, Profits > 0 (you can also write your own formulas)
- Top - filter the values based on a rank for example; top 10 or bottom 20