▶ See live dashboard on Tableau Public. No account required.
I created a Tableau dashboard showing the state of real estate property sales in Denmark since January 2020. The dashboard makes it possible to dig into data at postal code level.
The aim of this visualisation is to provide users with an overview of the market in a granular way that common listings websites don't provide (at the time I started the project at least).
I was looking to buy an apartment in 2020 and trying to get an overview of square meter prices by neighbourhood in Copenhagen. However, I could mostly find average prices at communal level. Copenhagen commune, as an example, is big (for Denmark) and prices vary a lot depending on neighbourhoods. Getting an average price per square meter for the Copenhagen commune is not useful for a prospective buyer - square meter prices can vary by 30-40.000 DKK for neighbourhoods barely 5 km from one another.
The project has 3 main parts:
- Web scraping of housing listings on boliga.dk, using Python, Pandas and the BeautifulSoup library
- Automation using Windows Task Scheduler
- Data visualisation and dashboard in Tableau
- As of February 2023, data is only related to apartments ("lejlighed" and "villalejlighed"). Boliga.dk does not use weighted area (defined by Finanstilsynet) when calculated the square meter price of a house, which makes it impossible to compare two houses together. I chose therefore not to include houses data in the dataset.
- Data is not representative of the full housing sales picture in Denmark. Since data is scraped from a housing sales website, it only includes listings that are made public. A lot of properties are indeed sold without having been made public.
- There can be small discrepancies in the date a property is put on sale or sold/removed from the website, which are dependant on how quickly the website publishes listings.
- Automation via Windows Task Scheduler is practical but not ideal. I would like to avoid having to leave my computer on sleep mode for it to work. I am looking into ways to take the automation online.