Skip to content

Latest commit

 

History

History
70 lines (51 loc) · 3.43 KB

barcelona.md

File metadata and controls

70 lines (51 loc) · 3.43 KB
Raw

Barcelona Airbnb Exercise

  • Degree of Difficulty: *
  • Goal: find those blocks and neighbourhoods with the number of Airbnb houses/rooms.
  • Features Highlighted:
    • Data edition and filter:
      • SQL console.
      • Dynamic filtering by widgets.
      • Layer selector.
    • Data Visualization:
      • Change basemap.
      • Blending methods.
      • Style marker-fill by value (choropleth).
      • Custom legends.
    • Geospatial analysis:
      • Intersect second layer.
  • Datasests needed:
    • Barcelona Airbnb data: airbnb_barcelona. You can download it from here.
    • Barcelona building footprint and neighborhoods: barcelona_building_footprint and barris_barcelona can be found in CARTO Data Library.

Barcelona Airbnb dataset was downloaded from the Inside Airbnb website. In order to download them to your local machines, you are using CARTO SQL API behind the scenes.

Workflow

  • Create a map with airbnb_barcelona (from your local machine), and barcelona_building_footprint and barris_barcelona datasets (from the Data Library).
  • Change the map title and layer names.
  • Change basemap: click on "Positron" and then select "Dark Matter (Lite)".

basemap

  • Go to Map Options (the slider kind of button at the blue left sidebar), activate LAYER SELECTOR.
  • Go back to the main menu.

layer-selector

  • Style "Airbnb" layer. Set marker-width to 2, remove STROKE and set a BLENDING method such as multiply.
  • Apply the following SQL query to the "Airbnb" layer in order to get the prize as a numeric field:
SELECT 
  *, 
  substring(price FROM '[0-9]+')::numeric as price_num 
FROM 
  listings_barcelona
  • Add some widgets from "Airbnb" layer (price_num, property_type and number_of_reviews, for example). We will add widgets from the other two layers later. Edit them.
  • Add a Intersect and aggregate analysis to the "Blocks" layer, and "Neighborhoods". Set "Airbnb" as TARGET LAYER and COUNT as OPERATION.

intersect

  • Add widgets for n_distri and n_barri columns in "Neighborhoods" layer. Try different aggregations and see how it affects the ordering in the widget.
  • Add ref_cadast widget from "Blocks" layer. Use COUNT as aggregation operation.
  • Style "Blocks" and "Neighborhoods" as a choropleth. For each layer, go to the STYLE tab. Secondly, click on the polygon-fill FILL color and select BY VALUE. Select count_vals_density. Set the number of buckets, color scheme (but invert the colors! Lighter colors should show higher values in contrast to the dark basemap) and classification method as you wish.

We have selected count_vals_density instead of count_vals, because when we make a choropleth map with polygon layers we need to use data which is normalized, in this case by the area.

  • Edit the "Blocks" and "Neighborhoods" legends.
  • Add a custom legend for "Airbnb" layer. Go to LEGEND tab, then click on CUSTOM and type "House/room".

custom-legend

barcelona