Typical Scenarios of LOD Expressions in Tableau

A free video tutorial from R-Tutorials Training
Data Science Education
Rating: 4.4 out of 5Instructor rating
24 courses
283,290 students
Typical Scenarios of LOD Expressions - Summary

Learn more from the full course

Advanced Tableau - Level of Detail Expressions / LOD

Take Your Tableau Data Visualization and Data Analytics Skills to the Next Level and Design Custom Solutions with Ease

06:12:03 of on-demand video • Updated September 2019

Understanding LOD expressions and using them confidently
Performing calculations in Tableau that are at a different level of detail than the view
Analyzing and solving complex analytical challenges
Understanding the different levels of details of multivariate datasets
Cohort analysis
Market basket analysis
User retention analysis
Binning aggregates by dimensions
Proportional brushing
Relative comparison of values/ categories
Nesting LOD expressions
English [Auto]
After completing all those projects and exercises, I think it would be useful to summarize the material and name a couple of scenarios where Lod expressions come in handy. In the first part of this course, you learned about what the view load and the calculation level of detail are in the Tableau environment. Then we went through the three types of load expressions include exclude and fixed. Here. We discussed what the different scoping keywords do and also how they compare to each other. At this stage, you learned about the first and very basic case of using an expression, namely when you need the data to be aggregated at a different level of detail as the view Lod. In this case, we mainly compared a member against a whole category or an average against a total. We also compared Lod calculations against reference lines, how to get the same results and also what kind of advantages the analyst has when choosing an expression over a reference line. So this was the first scenario comparison of data aggregated at different levels of detail. Of course, the intro lectures were not the only case where we used Lod expressions for comparison. We actually did that throughout the whole course, but in combination with other methods. After those intro parts, we went on with intermediate and advanced examples. The first one of them was the topic of binning aggregates. Here you learned that the fixed expression is ideal to create bins of a measure by using categories. Those categories can be defined by a dimension or some conditional statement. What makes the fixed expression ideal for that task is that Miss One can be used both as a measure and a dimension. After converting the custom calculation to a dimension, it can be used for histogram like charts alongside with a measure. Basically the second scenario where you want to consider using an Led expression is when you want to bin aggregates by a given condition or a category using the same principles in a slightly different way. Allows the user to perform even cohort analysis in Tableau. After that, we took advantage of another quality of the fixed expression, namely that it bypasses the dimensional filters, particularly in the lectures on proportional brushing. We use that one extensively. This way we could build visualizations where some measures regarded the filters while other ones ignored them. So you want to consider using an Lod expression in cases where you want to apply a filter to only some fields in the view, or if you want to calculation to include data that has been filtered out from the view. It is basically two sides of the same quality. The last general scenario I want to mention here is to use an expression to resolve the error message cannot mix aggregate a non aggregate values. The fixed expression is an ideal tool for that because that one evaluates the aggregation at the row level. Therefore it can be used alongside with row level expressions, including exclude work like some kind of complementation to the few lod. Therefore, the work at the aggregate level, this leaves the fixed expression the only useful tool for resolving that error message. Of course, apart from restructuring the whole view. So as you can see, there are a bunch of different ways in how to utilize load calculations in your daily work. As I said in the intro, some analytical questions are easy to ask, but can be surprisingly hard to answer with a general toolset of Tableau. Those are questions addressing different detail levels of the data at the same time. Sometimes you can find some sort of a workaround, but the final outcome can still contain too much churn, which is rather to be avoided in today's data science. You might find that this course was rather heavy on the fixed expression, and we talked less about the other two scoping keywords. This is because the include and exclude expressions are very limited compared to fixed. The way in. How you can utilize them is much more narrow. And basically the same process over and over again include adds to the view load while exclude subtracts from the view load. Sort of like a one trick pony. The fixed expression, on the other hand, is much more flexible and with that one the analyst has more freedom to come up with custom solutions. As you probably noticed during the course, the fixed expression is more universal while the other two have their particular place in the workflow. All right. I hope you found this summary helpful. If you have any questions concerning this course, feel free to ask them in the Q&A section of the course. Do also not forget to leave a rating and drop some lines on what you think about the course. Other than that, I wish you best luck in your life and career, and I hope that I'll see you again in one of my other courses.