Video

# Dates Magic – Analyzing Trends & Seasonality in Tableau

One of the first things that I was very impressed with Tableau when I tried it in early 2013  ago was the ease with which dates analysis can be done in Tableau. It felt like sheer magic to me then (and even now) – I came to Tableau from Powerpivot and Qlikview background. In those tools, you have to do a lot of work to enable just basic quarterly, monthly analysis. In Tableau, everything was just built-in. I just couldn’t believe that it was so easy.

In my opinion, this is one of the most valuable functionality for a typical business user. I have seen a client of mine, who is a owner of business, create many intricate graphs to analyze trends & seasonality with just an hour of training in Tableau  – in this video, I will show you the different options for understanding trend & seasonality in Tableau in 13 minutes.

Caveat: Now that I am a Tableau power user, there are a few quirks with how Tableau handles dates that can frustrate a new user – especially the discrete vs continuous part. But hey, we live in a world, where everything created by humans is imperfect in some way – so, let us enjoy what is available and be grateful 🙂

Video

# Golden Rule of Aggregation in Tableau

In this video, I talk about a simple rule that can be of immense help in understanding how advanced calculations work in Tableau – if you properly understand the Golden Rule, you are well on your own way to figuring out how any advanced table calculation or LOD calculation actually produces the result.

Here is the gist of the video:

Before you understand the Golden Rule, you need to understand one more fundamental idea that is crucial in the reporting world: Level of Detail or granularity.

We can understand Level of Detail of a report by comparing it to the level of detail we see on a Google Map as we zoom in and out. As we zoom in, we see more detail and as we zoom out, we see less detail – the same thing applies to a report/chart. As we add more dimensions, we see more detail and as we remove dimension, we see less detail. Hence, Level of Detail is nothing but the combination of dimensions on the report.

But what does a Dimension mean for the database? What does the database do when you add one more Dimension to your report?

For the database, Dimension are equivalent to Filters. As you add a Dimension, the database adds a Filter. If you have 3 dimensions on your report, then the database has to add  filters on each of the 3 dimensions to calculate the value of a cell in your report.

Hence, here is the Golden Rule: For each cell/mark,

1. Find the Filters i.e. Level of Detail
2. Apply the Filters
3. Then do the Calculation e.g. SUM

In one-line, it can be written as “Filters THEN Calculation

From this video onwards, I am not going to be producing a text version immediately. It is difficult for me to find large chunks of time to produce the videos and to write the text version. But as I find time, I will come back and update the post to have the text version – for now it is only going to be videos. I hope you find them to be of value.

Video

# Basics of Aggregation Part 2 – How do we make sense of Text?

It has been a long time since I have posted here – 4+ months. Number of external events happened that really threw my schedule into a disarray – the first and the biggest one of them was a cyclone that shook our city, Chennai (India), on 12 December 2016. This resulted in internet getting disconnected for a number of weeks and even when the connection came back, it was quite unstable. Then, we had a number of different political events happen which disrupted normal life. Hence, the flow I had in creating this series was disrupted. Once I have lost the flow, it was hard to come back. And project deadlines and international travel also made it difficult for me to get back to doing what I love.

Now, I had to spend a number of days trying to read/watch all that I have created to reorient myself. I really struggled to get started with the same energy and the ideas I had before. If any of you had read the book – Big Magic, you can understand the phenomenon I am going through better – once you start any creative endeavour, it is hard to resume if you take a break. Quite literally, the ideas go somewhere else and it is hard to bring them back. If you have not read the book, I suggest that you at least take a look at Ken Black’s short post where he talks about this book. By the way, Ken was the one who introduced this book to me and it helped me in so many ways to reflect consciously on the process of creating something – Thank you so much Ken.

Enough Stories – let us get back to the subject matter at hand. In this post, let us answer the 2nd of the following 2 questions that I introduced in the last post:

1. How do we make sense of Numbers?
2. How do we make sense of Text?

The last post answered the first question and this post will answer the second question.

Here is the gist of the video:

There are 3 ways to make sense of Text:

1. List the unique values
2. Count the number of unique values
3. Count the number of records.

When we are generating a report using any text value (which are called Dimensions in Tableau), then the default behaviour is to list the unique values. Counting the number of unique values (technically called as Distinct Count) is another thing we do often.

Here is a text-version of the video for those of you who prefer to read than to watch – please note that this is NOT a transcript of the video.

Video

# Basics of Aggregation Part 1 – How do we make sense of Numbers?

In this post, My aim is to answer the question ‘What really is an aggregation and why do we need it?’. The way I am going to explain aggregation is by helping you answer the following 2 questions:

1. How do we make sense of Numbers?
2. How do we make sense of Text?

Here is the gist of the video:

Aggregation is nothing a but a technical name to a highly intutive process that we all undertake when we are presented with a lot of information – when we have too much information, we tend to summarize it. So, that is what we are doing with numbers here – we are summarizing them to 4 typical summary values (Total, Min, Max & Average) to make sense of them without going through every single value.

Here is a text-version of the video for those of you who prefer to read than to watch – please note that this is NOT a transcript of the video.

Video

# Dimensions & Measures – Simplest explanation you have ever heard

It has been a while since I have posted on Table Calculations. After finishing the Paradoxical Problem 2 video, I knew that I had a lot of work to do on Paradoxical Problem 3 – Though I have presented the Paradoxical Problem 3 number of times, I felt that I still need to simplify the presentation more so that everyone can relate to my explanation easily. It is a challenge as the Paradoxical Problem 3 deals with a problem that is complex in itself. So, I decided to create a few videos on the ‘Basics of Data Analysis’ that would set up the stage for me to explain the Paradoxical Problem 3 effectively.

Thanks for all of you who have reached out to me asking when I would post the Paradoxical Problem. I was quite surprised to see a number of people who are not Tableau users are watching these videos. And I was also surprised to see that some very senior people in my network (Director and CXO-level resources) have watched the videos. This confirms my feeling that the problems I described are not limited to Tableau.

So, here is the first video in the ‘Basics of Data Analysis’ areas. Here my aim is to explain Dimensions and Measures in a simple way. I have heard a lot of different explanations of these 2 terms that are filled with jargons. I wanted to create an explanation that a normal business user can easily understand and I hope I have succeeded.

Here is the gist of the video:

Measures are nothing but Numbers

Dimensions are nothing but Text (+ Date)

Video

This is the third video in the series – Demystifying table calculations in Tableau.

Here, we will look at the second paradoxical problem. If you have not gone through the first paradoxical problem, I recommend you to do it first before proceeding.

Here is the gist of the video:

In the second paradox, we look at another very common business requirement in reporting : looking at figures cumulatively like Year-to-Date (YTD), Month-to-Date (MTD), etc. We have already established that accessing a previous record is difficult for the databases. This paradox also has the same issue, but it adds one more nuance to it.

Here is a text-version of the video for those of you who prefer to read than to watch – please note that this is NOT a transcript of the video.

Video