AIStk

backends/dask_backend.py — Dask Backend

Purpose

Enable out-of-core trajectory statistics on large AIS datasets using Dask.

Responsibilities

• Read large CSV/Parquet files with Dask DataFrame.
• Partition workload across cores/workers.
• Compute per-MMSI metrics: points, distance, straight-line distance, tortuosity, turn index, average/max SOG.

Usage Example

import dask.dataframe as dd
from aisdataset.backends.dask_backend import compute_stats_dask

ddf = dd.read_parquet("lake/clean/2024/*.parquet")
out = compute_stats_dask(ddf, level="mmsi")
print(out.head())

This site is open source. Improve this page.