Quick Start¶

This guide will get you up and running with TOAD in just a few minutes. For more detailed examples and explanations, see the Tutorials section.

Minimal Example¶

Here’s a simple example that demonstrates the core TOAD workflow:

from toad import TOAD
from toad.shifts import ASDETECT
from sklearn.cluster import HDBSCAN

# Initialize TOAD object with your data file
td = TOAD("data.nc")

# Detect abrupt shifts using ASDETECT method
td.compute_shifts("tas", method=ASDETECT())

# Cluster detected shifts using HDBSCAN
td.compute_clusters(
    var="tas",
    method=HDBSCAN(min_cluster_size=10),
)

# Visualize the results
td.plot.overview("tas")

Next Steps¶

Learn the basics: Check out the TOAD Basics tutorial for a comprehensive introduction
Customize methods: Learn how to Implement your own clustering method and Implement your own shifts detection method

Data Format¶

TOAD expects input data as:

NetCDF files (.nc) readable by xarray
xarray Dataset or DataArray objects
Data structured as 3D arrays: space × space × time

The time dimension can represent actual time or any other forcing variable or bifurcation parameter.

For more information about data requirements and formats, see the TOAD Basics tutorial.

Installation

Tutorials