Exercise 3

Contents

Exercise 3#

Part 1: Weekly reflection (1 point)

Write a short reflection of what you learned from this week’s lecture and discussion and how you are going affect your final project work. How will you take your learnings from this week into consideration in your final work?

Part 2: Think Beyond Maps (1 points)

This week we saw that maps are not the only way to visualize geographical data. Can you think of other non-cartographic methods to represent geographical information? Consider how these alternatives might offer unique insights compared to traditional maps.

Come up with an example different from those discussed in the tutorial: Identify a type of data that you would visualize without using a map. Describe how you would visualize this data, detailing the advantages of your chosen approach. Consider whether your method is more effective than a map in conveying the message. Write approximately half a page exploring this idea, being specific with your data and possibly selecting a real case for your discussion. You do not need to create a visualization but you are welcome to do so.

Part 3: A mapping challenge (8 points)

Choose two of the following three options. When making your map, pay attention to things we have learned so far. As usual, the aim is to make a map that isinformative,correct, andvisually pleasing.

Option A: A hexbin map of Big data

Use internet speed data from Ookla, first quarter of 2022, and visualize one of the columns you find interesting (average download speed, upload speed, latency etc.). Use a static Hexbin map for visualization and map the average value of the column in the hexagons. Feel free to be creative and also use other methods to visualize your data. You can try other vsiulization approaches if you can justify your choice.

Data can be downloaded from Ookla in parquet format using the following link:

https://ookla-open-data.s3.us-west-2.amazonaws.com/parquet/performance/type=mobile/year=2022/quarter=1/2022-01-01_performance_mobile_tiles.parquet

The data becomes very large in CSV format. However, if you need it in CSV format (for example for easier handling in QGIS), you can convert the data to CSV using Pandas library:

import pandas as pd

# Load your .parquet file
df = pd.read_parquet("https://ookla-open-data.s3.us-west-2.amazonaws.com/parquet/performance/type=mobile/year=2022/quarter=1/2022-01-01_performance_mobile_tiles.parquet")

# Save it as a CSV
df.to_csv('2022-01-01_performance_mobile_tiles.csv', index=False)

Alternatively, you can download the CSV version of data from here.

You can choose theextentof your map (e.g., Global, Europe). Specially if you are doing this in QGIS, working with the full data can be heavy so you may want to work on a smaller area.

Hexbin map in QGIS:

This week, we learned in the tutorials how to create hexbin maps in Python. If you want to create this kind of map using QGIS, there are a few different tools:

Option B: Mapping OD connections

The Origin-Destination lines tutorial introduced two ways to visualize flows – connections and the magnitudes of mobilities between places. If you want to dig deeper into these concepts, explore and visualize Origin-Destination dataset. An example of such data is City bike trips in Helsinki. Feel free to use other datasets that you find interesting. You are free to filter the data as you see fit to tell an effective story.

  • NB! You will need to connect those bike stations to locations with this data and connect the ODs similarly to what was done Week 2’s global static map with airports and route connections.

Edge-bundling tool

For enhanced visualization of complex flows, consider exploring other specialized tools such as the Edge-bundling tool for regional mobility flow data.

Developed by the Digital Geography Lab at the University of Helsinki, this Python-based tool uses edge-path bundling techniques to significantly reduce visual clutter in flow maps.

Flow maps in Python:

For a proportional line symbol map, you can scale the values and provide them to geopandas.plot with the parameter linewidth. Edgebundling might require extra effort, but there are a few examples in this repository.

Option C: Cartograms

Create an area cartogram. Feel free to use whichever regions you wish. For example, up-to-date population data in various administrative levels of Finland is available via Statistics Finland’s WFS and global population data by country is available, for example, through Natural Earth vector map. Visualize a variable on the map after distorting it.

Worldmapper and Datavizproject may provide some inspiration.

Submission:#

Please submit your report (including the map) as single pdf file to Moodle. You do not need to submit the code.

Deadline: Thursday 29.5.2025 - end of day

[ ]:
Contents