Dashboard Week, The Data Schools’ challenge to consultants nearing the end of their intense fourth month training before starting placement. The challenge for each day this week – create a dashboard and tell a story using a newly provided dataset each day.
Day 5: Melbourne Census of Land Use and Employment (CLUE)
The final challenge for the week saw us receive a collection of documents that together captured census data for the inner part of Melbourne – surrounding the CBD. The pressure point for this challenge was that we only had 6 hours before we needed to present our story dashboards.
The data was already cleaned, aggregated and ready to be looked at in Tableau. The information contained in the census focused on land use where the number of dwellings, cafes, restaurants, bars, and pubs where recorded. Each of these locations each had a longitude and latitude so using a map as well as minor details.
Approach/Story
My idea for the story of my dashboard was to play on the fact that Melbournians love their food and drinks. The idea being that I would create a ‘quality of life’ score based on the number of hospitality services within each block/suburb. Think of a map of suburbs and each one has a score to show which suburb has the better services available. Overall, this was a fun approach to the dataset played on the fact that Melbournians only care about food and drinks.
Although I had a clean idea for my story, my approach to the challenge was what let me down and created more challenges than there needed to be. This was mainly due the fact that I didn’t inspect all the data and realise I didn’t need to aggregate the data myself. I sadly didn’t realise the data was aggregated until one hour until the presentations and it was too late. I did manage to produce a dashboard but I wasn’t happy with the outcome – especially when we were presenting to the office staff.
Data preparation
Data preparation took up most of my time and the main task was to aggregate the data for each block and suburb. The challenge with aggregating the data was that I needed to combine and link multiple tables. While this wasn’t difficult to achieve, I did spend some time exploring different ideas. One of which was to find the distance from each dwelling to each hospitality service within the dataset – I quickly realised that this would take too long to process.
Here is a snippet of my messy workflow (It’s not usually so poorly documented)
Dashboard creation
The challenging part about creating my dashboard was trying to create it given the limited amount of time I had allowed myself. In addition, I also encountered a few minor challenges, but this was eating up time.
Quality of life in Melbourne (link to dashboard)
Functionality of the dashboard allows users to see the quality-of-life score for each suburb and then drill down for further analysis. Within each drill-down, users can see the hospitality services on the map and see how the number of services compare to other suburbs.
While I have published this dashboard on Tableau Public, I will revisit this challenge and update my dashboard to create something that I am happy with!
