An easy quick-start guide to CBS open data

/ Author: Miriam van der Sangen
© Sjoerd van der Hucht Fotografie
In 2014, Statistics Netherlands (CBS) rolled out a portal to make its open data more publicly available. Since then, the use of open data has grown explosively. More and more users, ranging from public authorities to businesses and private citizens, are finding their way to the portal. In a bid to lend a further boost to the use of open data, CBS made available a number of user-friendly instruction guides on its website that enable users to get started quickly and easily with Python or R as programming languages.

Greater transparency

The open data portal was launched in June 2014. CBS open data coordinator Erwin van Mierlo explains: ‘The introduction was in response to an increased demand in society for greater transparency. CBS’ statistical information has become better accessible as a result of this portal and re-using has been made easier. We make sure the latest information is always available, also through our open data portal. This is a major advantage.’ Every month, on average 450,000 unique users make use of the datasets in the CBS open data portal. ‘Users from the public and the private sector are developing more and more applications which leverage the added value of our open data,’ Van Mierlo adds.

Quick and easy start

In order to make use of CBS’ open data sets, a so-called application programming interface (API) has been provided. A new version of this API is currently in development which is based on the OData 4-protocol. This led Jolien Oomens, ICT researcher at CBS, to write a number of instruction guides. ‘These will allow new users to get started in Python and R quickly and easily and to produce their own beautiful maps and time series plots. The package for R as mentioned in the guides has been developed by CBS methodologist Edwin de Jonge. The Python package has been developed by Jonathan de Bruin, Research Data Engineer at Utrecht University.

Lowering the threshold

Oomens designs and teaches courses within the CBS organisation that cover open data, deep learning, visualisations and R. She also develops the learning materials. ‘Instruction guides for OData 4 with Python and R have been tested by teachers and students from several universities and universities of applied sciences. We then processed their feedback. Some teachers indicated that they would like to introduce these guides at the start of the next academic year. This is why we decided to adapt the materials to the already existing OData 3 protocol and to publish them on our website as quickly as possible.’ Oomens hopes that the instructions in both Dutch and English will lower the threshold for users to start using open data.

‘We see more and more people - data analysts, data journalists and students, for example –wanting to work in a data-driven way. Hopefully these guides will make CBS data even better accessible.’