- Data is stored at tresorit; check your e-mail for the link
- The data is saved as parquet files
- The schema of the datasets provided is (convert to Danish time + categoricals if you like):
Schema([('MMSI', String),
('# Timestamp', Datetime(time_unit='us', time_zone=None)),
('Type of mobile', String),
('Latitude', Float64),
('Longitude', Float64),
('Navigational status', String),
('ROT', Float64),
('SOG', Float64),
('COG', Float64),
('Heading', Float64),
('IMO', String),
('Callsign', String),
('Name', String),
('Ship type', String),
('Cargo type', String),
('Width', Float64),
('Length', Float64),
('Type of position fixing device', String),
('Draught', Float64),
('Destination', String),
('ETA', Datetime(time_unit='us', time_zone=None)),
('Data source type', String),
('A', Float64),
('B', Float64),
('C', Float64),
('D', Float64)])- Slides
- See also the original documentation.
- The data has been transformed from the original zip files to the parquet files stored on tresorit using the script here
- See script
- See script
- See script
- See script
Every hour:
Every 15 minutes:
- See script
uv venv
# On Windows replace with: .venv\Scripts\activate
source .venv/bin/activate
uv pip install -e .