Are you familiar with the OpenStreetMap Statistics created by Piebro? It’s a handy tool that analyzes OSM changeset files and creates graphs from various perspectives.

• https://piebro.github.io/openstreetmap-statistics/
• https://github.com/piebro/openstreetmap-statistics

The project’s README also indicates that analysis can be performed for individual countries or regions. I analysed the Japanese region in my local environment, so I’ll summarize the procedure.

My Environment

• MacOS
• Python 3.9

Data Preparation

Clone Piebro's git repository to your local machine:
$ git clone https://github.com/piebro/openstreetmap-statistics.git
$ cd openstreetmap-statistics

Install osmium-tool using Homebrew:
$ brew install osmium-tool

Prepare the virtual environment:
$ python3 -m venv .venv

Install necessary dependencies:
$ pip3 install -r requirement.txt

Obtaining the OSM Changeset File

Piebro’s README uses torrent. Using torrent should result in lower network load, so it’s preferable if possible.

$ rm $(ls *.osm.bz2)
$ wget -N https://planet.openstreetmap.org/planet/changesets-latest.osm.bz2

Obtain the Poly File for the Target Region

Download the poly file for the target region from the Geofabrik site.

$ wget -N https://download.geofabrik.de/asia/japan.poly

Extracting the Changeset File

Use a poly file to extract the target region from the changeset-latest file. Please use the following script for extraction (written by Claude):

https://gist.github.com/nyampire/57360f29be2bff4d74d94e3c5bfa3237

Install shapely 1.8.0.
The latest version 2.0 and later seems incompatible with dependent software.
$ pip3 install shapely==1.8.0
$ wget -N https://gist.githubusercontent.com/nyampire/57360f29be2bff4d74d94e3c5bfa3237/raw/1fa9ded4126b3282af273469225407847cb38ee7/parse-latest.py
$ python3 parse-latest.py [poly.filename] [input.filename] [output.filename]
$ (sample) python3 parse-latest.py japan.poly changesets-latest.osm.bz2 japan-chansegets.osm.bz2

As a result of the script processing, in this example, japan-changesets.osm.bz2 will be output.

Generating Files for the Notebook

The subsequent processing is similar to Piebro’s procedure. I omitted the pv command as it’s only used to monitor progress.

$ rm -r -d temp
$ osmium cat --output-format opl output.osm.bz2 | python3 src/changeset_to_parquet.py temp
$ python3 src/parquet_to_json_stats.py temp

Updating the Notebook

Execute the following in the terminal. You can make it into a shell script and kick it, but it will also work if you paste it directly into the shell.

for notebook in $(find src/questions -name calculations.ipynb); do
    jupyter nbconvert --to notebook --execute "$notebook" --output calculations.ipynb
done

Starting the Local Server

Start a local http server to check. You need to start an http server; otherwise, the graphs won’t display due to CORS issues.

python -m http.server 1010

Access to “http://localhost:1010” to show graphs.

Instead of starting an http server, you should also be able to push to your personal git repository and check from gh-pages.

However, with the steps so far, the remote is set to Piebro’s repository. Please modify it to upload to your own repository.