Skip to content

Latest commit

 

History

History
39 lines (26 loc) · 3.83 KB

README.md

File metadata and controls

39 lines (26 loc) · 3.83 KB
Raw

sweden-rent-dashboard

An interactive dashboard built with python that enables you to visualise how rent prices differ across Sweden.

The dashboard/web-app generated from this project can be viewed by clicking here The dashboard was built to be highly interactive so please do feel free to interact with the figures, dialog buttons, sliders and text inputs. (Unfortunately, the web-app does not render well on mobile devices.)

Preview of the Dashboard Overview Page

SwedenDashBoard

Code and Resources Used

  • Python Version: 3.8
  • Packages Used: pandas, numpy, json, dash, plotly, BeautifulSoup, requests, urllib.request (Those that are required for hosting the web-app can be installed using: pip install -r requirements.txt after cloning the repo).
  • Statistical Data: Downloaded from Statistics Sweden. The median values were used instead of the means as the underlying distribution of each data set is unavailable.
  • Web scraping: Performed on various sites including wiki, Information Sverige and Open Street Map.
  • GeoJSON (Map) Data: Obtained from Open Street Map using their API. See section below for further details.

Repository Layout

Main folder

  • app.py: Used to generate the Plotly/Dash web-app. In order to run this you will need to have run all the below scripts in advance or use the files provided in the "assets" folder in this repo.

  • "get_kommun_county_info.py": This script web scrapes from both wiki and Information Sverige to:

    • Generate a dictionary that states what county ("län" in Swedish) each municipality ("kommun" in Swedish) belongs to.
    • Obtain a short bit of introductory text about each municipality.
    • Store the web address for each municipality's page on Information Sverige.

  • "prepare_rent_data.py": Takes the 4 Statistics Sweden excel files (these can be found in the "stats" folder) and cleans/reformats them. Outputs are saved as ".csv" files in the "assets" folder and are loaded into the web-app.

  • "get_geojson_data.py": This script first web scrapes relation numbers (like an i.d. number for a map file) for all counties ("län" in Swedish) and municipalities ("kommuner" in Swedish) in Sweden from Open Street Map. The relation numbers are then used to download GeoJSON files from the OSM database and merged to create maps of Sweden (with borders marked at both the county and municipality levels). The original map files generated from this process ("counties_map.json" and "kommuner_map.json") and they were then "simplified" (resolution decreased) using mapshaper to improve page loading times on the web-app.

Folder: stats

The four ".xlsx" files were obtained directly from Statistics Sweden and left unaltered. The file "sources.txt" provides additional information about how exactly these files were obtained.

Folder: assets

These are the resources read in and used by the Plotly/Dash web-app. These were all generated in advance using the scripts described in the main folder.

Issues/Comments/Questions

Please feel free to open an issue or pull request if you have any issues/comments/questions or notice something that could be improved.