In this Tableau data visualization tutorial, we’ll learn to use the LOOKUP table calculation function to return sales revenue for the same day last year. A number of different techniques are used in the creation of this dashboard.

I designed this dashboard solely as a teaching exercise to help you understand the LOOKUP function and how to show the same date last year in a separate column.

As we learned in a previous video Tableau Table Calculations Simplified, (make sure to watch this video afterwards for more clarity), we’ll compute using specific dimensions and then use “At the level” to make sure our LOOKUP table calculation is performing correctly.

The “Show Missing Values” option is selected to fill in date gaps in the data set that do not exist. Ensuring 365 dates per year are present in the visualization enables the offset (i.e., -1) in the LOOKUP calculation to arrive at correct sales revenue from the same day in the previous year.

You’ll learn that we can filter on a table calculation by using another table calculation. Filters based on table calculations do not filter out underlying data. Instead, the data is hidden from the view, allowing dimension members to be hidden from the view without impacting the data in the view.

Observe the Tableau filter order of operations above. Applying a dimension filter before the Table Calculation filter removes underlying data which affects the proper functioning of Table calculations. Typically, Table Calculations only work on values that are visible in the view. By applying a table calculation (which is last in the order of operations) you preserve underlying data but filter out data from the view.

Interact with this dashboard via the picture link:

You need to read these posts and watch these videos for additional information:

In this Tableau data visualization tutorial I used a technique shared by Tableau Zen Master Ryan Sleeper to “equalize” dates across the same axis. This date equalization calculated field enables year over year, quarter over quarter, month over month, week over week and same day last year comparisons.

It’s a pretty clever way of preserving the same month and day of date values across many years and updating their respective years to one common year.

For example, all dates would retain their current month and day but would share the year value of ‘2018’. In this manner, data points from various years can be stacked on top of each other for comparison purposes.

Additionally, by creating a parameter value for a specific date part (i.e., year, month, week, etc.,) the user has control over the level of comparisons in the visualization.

You’ll have to watch the video to see the details. Again, thanks to Ryan Sleeper for sharing this tip with the Tableau community which enabled me to apply it to my dataset and share it with you in video form.

As always, If you find this type of instruction valuable make sure to subscribe to my Youtube channel.

All views and opinions are solely my own and do NOT necessarily reflect those my employer.

In this Tableau tutorial I will discuss a basic quick table calculation and try to demystify what is happening behind the scenes. All of the following will hopefully be made more clear in the video but I’m sharing the text below for reference after you watch the video.

I can’t take credit for the content in this post as the Tableau online help site has some quality information on table calculations which I will reference below. There are two important concepts in understanding table calculations: partitioning and addressing.

Key Concepts with Table Calculations: Addressing vs. Partitioning

The dimensions that define how to group the calculation (the scope of data it is performed on) are called partitioning fields. The table calculation is performed separately within each partition.

Partitioning fields break the view up into multiple sub-views (or sub-tables), and then the table calculation is applied to the marks within each such partition.

The remaining dimensions, upon which the table calculation is performed, are called addressing fields, and determine the direction of the calculation.

The direction in which the calculation moves (for example, in calculating a running sum, or computing the difference between values) is determined by the addressing fields.

So when you order the fields in the Specific Dimensions section of the Table Calculation dialog box from top to bottom, you are specifying the direction in which the calculation moves through the various marks in the partition.

When you add a table calculation using the Compute Using options, Tableau identifies some dimensions as addressing and others as partitioning automatically, as a result of your selections.

But when you use Specific Dimensions, then it’s up to you to determine which dimensions are for addressing and which for partitioning.

At the level (Partitioning)

The At the level option is only available when you select Specific Dimensions in the Table Calculations dialog box, and when more than one dimension is selected in the field immediately below the Compute Using options —that is, when more than one dimension is defined as an addressing field.

This option is not available when you’re defining a table calculation with Compute Using, because those values establish partitions by position. But with Specific Dimensions, because the visual structure and the table calculation are not necessarily aligned, the At the level option is available to let you fine-tune your calculation.

Use this setting to set a break (that is, restart of the calculation) in the view, based on a particular dimension. How is this different from just using that dimension for partitioning? In fact, it is partitioning, but it’s partitioning by position rather than by value, which is how partitioning is defined with the Compute Using options.

Filtering on Table Calculations in Tableau

Filtering on Table Calculations in Tableau can be a bit hacky. More often than not, table calculations are dependent upon the data in the view. That means in order to filter on a table calculation, we need a way to preserve underlying data and only hide data from the view.

Filters based on table calculations do not filter out underlying data in the data set, because table calculation filters are applied last in the order of operations. Instead, the data is hidden from the view, allowing dimension members to be hidden from the view without impacting the data in the view.

Notice in the order of operations diagram below how Dimension and Measure filters are applied before Table calculation filters. When trying to filter on a table calculation (which is dependent upon the data in the view) the results may be unexpected. If we turn our dimension or measure into a table calculation, we can then filter the Table calculation at the appropriate level, which preserves underlying data and only hides the table calculation values we wish to filter from the view.

The content in this post was quoted from the sources below.

As always, If you find this type of instruction valuable make sure to subscribe to my Youtube channel.

I want you to increase your efficiency and to stop using spreadsheets for every single analysis.

Everybody works with time series data at some point in time. Year over year (also known as YoY) analysis is one of the most useful analyses you can perform to determine changes, analyze growth and recognize trends in quantity on an annual basis.

Unfortunately, most data preparers are used to performing some unaesthetic flavor of this analysis using only Excel (looking at you FP&A). Without the benefit of using visualization to easily recognize trends, data consumers are forced to work harder to tease out the most salient information.

If you have access to Power BI Desktop (available for free), then you can perform a tabular year over year difference calculation and then tie that information to a bar chart that will help you visualize the variances.

In this video I will show you how to create a calendar table in DAX (Microsoft’s formula expression language) and use that table to enable a year over year analysis of customer orders at fictional Stark Industries. You don’t need to be an expert in DAX to take advantage, just type in the date calendar formula you see in the video and tweak the simple calculations to fit your data.

You could obviously perform a simple YoY analysis in Excel, but I want you to stay relevant and learn something new!

If you find this type of instruction valuable make sure to subscribe to my Youtube channel.

In this video tutorial I describe a dashboard that I put together that displays the distribution of various NBA player statistics. I use the always handy parameter to enable the user to choose which statistics are displayed on the dashboard. Although I’m showing sports statistics measures in this dashboard, it could easily be repurposed to show the distribution of a variety of business related metrics.

I break the dashboard up into three areas: histogram, dot strip plot, and heat map. In the second part of the video, I describe in detail how to build out a jittered dot strip plot. The benefit of the jittered dot strip plot is that the marks representing NBA players obstruct each other much less as compared to the linear dot strip plot.

Feel free to head to my Tableau Public page and download the workbook for yourself. Drop me a line in the comments or on YouTube if you learned something.

As always, do great things with your data!

If you find this type of instruction valuable make sure to subscribe to my Youtube channel.

In this post you’re getting two videos for the price of one (considering they’re all free for now, that’s a good thing). I put together a relatively simple dashboard to help illustrate a few intermediate level concepts. In this first video I take a look at the number of total assists by NBA players during the 2017-2018 season. In case you were wondering, Russell Westbrook led the league in assists during that season. If you don’t know who Russell Westbrook is, then skip this Tableau stuff and watch the last video immediately (and then come back to the Tableau stuff).

In the first Tableau dashboard video, you’ll learn two concepts:

How to make rounded bar charts;

How to filter the number of bar chart marks via use of a parameter;

Part 2: Apply Custom Sorting in Tableau

In the second video I build upon the dashboard built in the first video by showing you how to add a custom sort. The custom sort relies upon the creation of a parameter and a calculated field. The parameter and calculated field enable the user to select either a dimension (e.g., Player Name) or a measure (e.g., sum of assists) from a drop down box and the visualization will sort ascending or descending as requested.

The calculated field relies upon the RANK_UNIQUE function.

In this context, RANK_UNIQUE returns the unique rank of each player’s assist total. The key with RANK_UNIQUE is that identical values are assigned different ranks. As an example, the set of values (6, 9, 9, 14) would be ranked (4, 2, 3, 1), as no tied rankings are allowed.

Part 3: Interact with the Dashboard

Bonus: Russell Westbrook on the Attack

For those of you who do not know who Russell Westbrook is, I’ve got you covered. These aren’t assists but in these situations, he didn’t need to pass!

References:

Thanks to both the Tableau Magic blog for outlining the concept of rounded bar charts and the VizJockey blog for the custom sort methodology. Check out and support these blogs!

As always, do great things with your data!

If you find this type of instruction valuable make sure to subscribe to my Youtube channel.

In this video I tackle the subject of slope graphs also known as slope charts. I had some fun putting together this dashboard that illustrates the changes in wins for NBA teams during the 2016-2017 and 2017-2018 seasons. From the video you’ll discover that Chicago, Atlanta and Memphis are on a Hindenburg-like trajectory, while trusting the process in Philadelphia led to huge season gains in overall wins.

Here’s what you will learn from this video:

How to create a parameter that enables a user to select which win statistic measure to visualize;

How to use Table calculations like, LOOKUP(), FIRST() and LAST() to calculate period over period change;

How the impact of Mike Conley’s injury affected the Memphis Grizzlies last season;

Click the pic to interact with the Tableau Public visualization, also download the workbook and data to dissect as needed.

For your convenience the calculated fields that I used to create the measures are listed here. Note that [Selected Measure] is a parameter that you need to create that lists all of the measures.

Calc Select Measure CASE [Selected Measure] WHEN “Home Losses” Then [Home Losses] WHEN “Home Wins” Then [Home Wins] WHEN “Overall Losses” Then [Overall Losses] WHEN “Overall Wins” Then [Overall Wins] WHEN “Road Losses” Then [Road Losses] WHEN “Road Wins” Then [Road Wins] WHEN “vs East Conf Losses” Then [vs East Conf Losses] WHEN “vs East Conf Wins” Then [vs East Conf Wins] WHEN “vs West Conf Losses” Then [vs West Conf Losses] WHEN “vs West Conf Wins” Then [vs West Conf Wins]

END Better or Worse IF [Selected Measure] = “Home Wins” OR [Selected Measure] = “Overall Wins” OR [Selected Measure] = “Road Wins” OR [Selected Measure] = “vs East Conf Wins” OR [Selected Measure] = “vs West Conf Wins” THEN //WIN MEASURES: Negative delta treated as “WORSE”, Positive delta treated as “BETTER” (IF [Delta] < 0 THEN “WORSE” ELSEIF [Delta] = 0 THEN “SAME” ELSE “BETTER” END) ELSE //LOSS MEASURES: Positive delta treated as “WORSE” (more losses are worse), Negative delta treated as “BETTER” (IF [Delta] > 0 THEN “WORSE” ELSEIF [Delta] = 0 THEN “SAME” ELSE “BETTER” END) END Delta LOOKUP(SUM([Calc Select Measure]),LAST()) – LOOKUP(SUM([Calc Select Measure]),FIRST()) Delta ABS Value ABS(LOOKUP(SUM([Calc Select Measure]),LAST()) – LOOKUP(SUM([Calc Select Measure]),FIRST())) ToolTip <Team> Trend: <AGG(Better or Worse)> by <AGG(Delta ABS Value)> During the <Season> Season, the <Team> had <SUM(Calc Select Measure)> <Parameters.Selected Measure>.

I have to give thanks to Ben Jones at the Data Remixed blog for the inspiration!

As always, If you find this type of instruction valuable make sure to subscribe to my Youtube channel.

In this tip I will provide an overview of a completely free data profiling tool (at least as of the time of this post) that is easy to use if you are in the Microsoft database stack. You can download the IDERA SQL Profiling tool and immediately put it to work to perform basic column profiling and analyses. The tool will display a summary of the data contained in a selected table and each of its columns.

Use this tool on the following systems:

Microsoft SQL Server: 2008 R2, 2012, 2014, 2016; 2017 Windows & Linux (provisional); Express, Standard, Enterprise editions

Microsoft Azure SQL Database

Amazon Relational Database Service (RDS)

IDERA SQL Data Profiler has some minor quirks but you can’t beat the price. Check out my review in the video above.

Just remember that data profiling should always be done initially before you start analyzing a new dataset or designing a new visualization. Always start with the basics.

If you find this type of instruction valuable make sure to subscribe to my Youtube channel.

In this video you will learn how to filter the top N values shown in your bar chart visualization using a slicer.

This technique uses one measure that generates a number 1-10, that will be applied to a slicer.

Another measure will basically rank all of the values associated with your data bars and only return the values that are less than or equal to the number you select in the slicer.

The comments that I apply to the DAX function should help make it easy to understand. I have to give a shoutout to GilbertQ from the PowerBI community for coming up with the initial approach which I tweaked for the video.

As always, If you find this type of instruction valuable make sure to subscribe to my Youtube channel!

Forecasting in Tableau uses a technique known as exponential smoothing. This is when an algorithm tries to find a regular pattern in your data that can be continued into the future.

In this video I’ll share some helpful tips to help you determine which options you should select that will enable Tableau to make the most predictive forecast for your data. By the end of the video you will be able to differentiate between an additive and multiplicative data pattern and to evaluate MASE to measure the accuracy of the forecast.

I’m not talking about this Mase:

Rather, you’ll learn about the mean absolute scaled error (i.e., MASE) and how it helps you judge the quality of the model.

In addition, you’ll also also learn how to compare your actual data to the Tableau forecast in order to judge if the model is doing its job.

If you’ve used the forecasting capabilities in Tableau without knowing about these concepts, you might have generated an inaccurate error riddled forecast. Don’t just set a forecast and forget it. Watch this video and generate better forecasts in Tableau!

Here is additional reading from Tableau on the forecast descriptions (including MASE).

As always, If you find this type of instruction valuable make sure to subscribe to my Youtube channel.