Our Blog

where we write about the things we love

30

Mar

Datarama: visualising statistical data

I was lucky to join great minds at the Datarama hackathon, an exciting event organised by Stats NZ and partners (HackMiramar, Koordinates, data.gov.nz, nexus, Microsoft, Open Data NZ). The event was held on 24-25 March 2017 at the Mahuki Innovation Hub, Te Papa, Wellington.

The challenge was to define the use case (set the business or social task, who will benefit from the solution and how), select applicable data sets, suggest a solution, describe current barriers to use the data and areas of improvement.

The resulting projects could be used to guide Stats NZ in their current and future open data work. It also enabled participants to collaborate in a mini-project and share their experiences using the API/ statistical data.

Datarama Team 100% Tourism

Our awesome team (“100%? Tourism”) was made up of Keitha Booth, Nick Willetts, Ana-Maria Mocanu, Thomas Kernreiter, Andrew Sknar, Chris Auld, Adam Fabish and Eddie Samuel. We were a team of strangers that got together to apply different skill-sets to the project idea. 

Our working Proof Of Concept is a one stop shop based on regional data (local view), to access international tourist data, mainly geared towards small tourist operators. It supports forecasting, provides exportable files, and visualisation. Where data for each region needs to be exported separately we selected some Regional samples instead of full scope. In order to achieve a feasible result in the short timeframe, focus was narrowed-down to international visits to the North Island of New Zealand.

We used a staging storage applying filtered information and an open data source feed. We then used R and Python libraries to create queries and output a visual presentation. The core conceptual diagram the Solution is shown below:

Datarama Conceptual Diagram

The Team combined parameters for analysis with seasonality and visualisation. The outcome is viewable via applying parameters: Regional; Time range, type of visiting.

Trend in expenditures by regions is added to the analysis. Among our outcome observations: expenditures (while total value is up, the average pay-check is decreasing; even more drastically as data shown as nominal, inflation to be added); twist of factors can be visible around 2015; from another angle there was also growth in domestic tourist spend for the period 2009-2015 (incl. spatial data).

Out of eight teams, ours was one of the few to complete the task (working core solution) in the allocated time (one hour of idea refining plus 6.5 hours of actual project work, and two minutes of compressed presentation).

Following the presentations and project work assessment by jury, our team was awarded the main prize “Most potential to change lives”, presented by Chief Statistician, Liz MacPherson.

I am proud of our team where individually contributed efforts did not just simply added-up incrementally, but brought this mini-project to a new level of cohesive work helping Stats NZ (now including Open Data NZ) in building APIs and supporting transparency and efficiency of valuable data analysis. It is not just our team who should feel as winners, but New Zealand public and private communities who are benefiting from using Open statistical data.

Posted by: Andrew Sknar, Senior Consultant | 30 March 2017

Tags: Open Data, data insights, Hackaton


Top Rated Posts

Blog archive

Stay up to date with all insights from the Intergen blog