The challenge for dashboard week day 4 is using a very old version of Tableau from 2015, tableau v8.3. We also had to use the cocktail dataset from the cocktail database. The dataset itself was simple enough (but this often means not many insights). The real challenge is using Tableau v8.3 to build a dashboard.

Things that didn’t work in Tableau 8.3

I think Tableau has come a long way in the past 5 years, especially in the area of user interface and interactivity. This experience has really made me appreciate all the improvements made throughout the years. Things that didn’t work back in version 8.3 include:

  • Level of Details: I didn’t really need this for this dashboard but not having a level of details would be really inconvenient for a complicated dataset
  • Transparent Background: all the pretty designs we see right now on Tableau public will be harder to make without the transparent background. You have match the background colour with the picture background instead, which was inconvenient.
  • Set Actions: I once watched a video by Bethony Lyon when Set Actions were first released and she was so passionate talking about it. I kind of get it now.
  • Suggestions in Calculated Field: in the old version, you have to search for the field and double click on it in the calculated field. As a result, if you don’t remember exactly your field name, the calculation may not work. In the current version, Tableau automatically suggests the field name which makes creating calculated fields so much easier.
  • Outdated ‘look’: this may just be personal taste but compared to the sleek design of Tableau now, v8.3’s user interface looks outdated.

Step 1: Prepping data using Alteryx

As usual, I began my analysis with prepping data using Alteryx. I did the following things to prep:

(1) Concatenate ingredients & measures into 1 field and break each ingredient down into different lines
(2) Create new measures such as # of alcohol shots, level of difficulty (based on how long the instruction is)

I also wasted a lot of time because I couldn’t decide whether I wanted the data to be tall or wide.

Step 2: Visualisation in Tableau (v8.3)

Here came the part where I was very intimidated with for this challenge. The interface of Tableau feels different and I didn’t feel very comfortable working in it, which sort of gets in the way of creativity. Since we didn’t have much time yesterday for familiarizing myself with the interface and trying out complex things, I decided to keep my dashboard simple. My topic was on mixing cocktails using different types of filters and parameters. I think it was a good opportunity for me to try out different types of filters such as a wild card match (which I rarely used in my dashboard). It turns out that lots of prepping work I did in Alteryx were somewhat unnecessary.

(1) Prepare design in Canva

Whenever I have a creativity block in Tableau (with the blank canvas), I’d go to Canva to create some nice background template. It sort of helps guiding me with my dashboard layout as well. I picked one of Canva’s existing black template with an alcohol theme using a custom dimension of 1000 x 800 (tableau’s default dashboard size). I needed to go back and forth a little to adjust the design in Canva. Some of the design elements get in the way of the dashboard.

(2) Try out different filters in Tableau

For my dashboard, I wanted to have different types of filters so that the users can nagivate to the cocktail of their choice, see its photo and instruction for making it. The default filter in Tableau would be radio box and drop-down which are a bit small to see and may be hard to use so I tried out other approaches with filters. Some of the things I tried include button filters, parameter drop-down, and wild cards.

Button Filter: This is just a label mark with filter action.
Parameter drop-down: This filter is a parameter with a list of liquor bases and a calculated field using CONTAIN() function. When selecting a liquor from the list, Tableau will look up that liquor in the list of ingredients.
Wild cards: I mostly used wild cards for searching a particular string (e.g restaurants, schools). I’ve never used it for one dimension to look up value in another dimension, so sometimes I forgot I can do that. It was a good opportunity to try that for this challenge where I can look up drinks using what’s in the ingredients. If you type ‘egg’, you’ll be able to look up drinks that contain eggs.

(3) Custom shape

I read about making a map with custom shape in Tableau before and decided to try out for this challenge. There was no map in the dataset so I used it for the photo of the cocktails.


The Data School
Author: The Data School