streets-monitoring

This GitHub repo documents changes between the yesterday and today streets and adresses downloaded from official STAC repo: https://data.geo.admin.ch/ch.swisstopo.amtliches-strassenverzeichnis/amtliches-strassenverzeichnis/.

Hardware requirements

No specific requirements.

Software Requirements

No specific requirements.

Folder structure

├── .github                       # actions directory
  └── workflows
      ├── actions.yml             # action downloading and processing the data
      └── actions_rm.yml          # action deleting old downlaoded data folders
├── data
   ├── YYYY-MM-DD                 # data folders
      ├── .gdb                    # GDB with geometries and ESID
      └── .csv                    # CSV with data attributes                     
  └── *output files               # output files after processing
├── main.py                       # processing scripts run during GitHub action
├── recap.csv                     # CSV with monitoring overview
├── requirements.txt              # libraries used by GitHub action to build the python environnement 
└── status.log                    # Stauts log from runs

Scripts and Procedure

The script makes a join between the GDBs of yesterday and today. The join is based on the geometries, which must be strictly identical to be joined. In recap.CSV, there is a summary of changes between the two dates:

• previous_date: yesterday
• new_date: today
• sum_ESID_diff: number of ESIDs that differ
• sum_LABEL_diff: number of different adress labels
• sum_previous_ESID: the number of geometries/streets/adresses that only exist in the previous_date GDB
• sum_new_ESID: the number of geometries/streets/adresses that only exist in the new_date GDB

Vector results in data folder:

• YYYY-MM-DD_duplicate.gpkg: geometries duplicated in the original GDBs. Only one geometry per duplicate is kept for further analysis.
• YYYY-MM-DD_YYYY-MM-DD_previous.gpkg: geometries/streets/adresses that exist only in the previous_date GDB.
• YYYY-MM-DD_YYYY-MM-DD_new.gpkg: geometries/streets/addresses that only exist in the new_date GDB.