Quickstart

Before you start, install the released package:

python -m pip install de-time

Contributor checkout:

python -m pip install -e .[dev,docs]

Python

import numpy as np

from detime import DecompositionConfig, decompose

t = np.arange(120, dtype=float)
series = 0.03 * t + np.sin(2.0 * np.pi * t / 12.0)

result = decompose(
    series,
    DecompositionConfig(
        method="SSA",
        params={"window": 24, "rank": 6, "primary_period": 12},
    ),
)

print(result.trend.shape)
print(result.meta["backend_used"])

Expected output on a native-enabled build:

(120,)
native

The result object carries the same shape contract for every method:

print(result.season.shape)
print(result.residual.shape)
print(sorted(result.meta))

(120,)
(120,)
['backend_requested', 'backend_used', 'input_shape', 'method', 'n_channels', ...]

CLI

detime run \
  --method STD \
  --series examples/data/example_series.csv \
  --col value \
  --param period=12 \
  --out_dir out/std_run \
  --output-mode summary

Expected stdout:

Running STD on examples/data/example_series.csv...
Done. Results saved to out/std_run

This writes out/std_run/example_series_summary.json. The same command with --plot also writes decomposition and residual figures.

For a published plot-backed run, see the Univariate Tutorial.

Multivariate

import numpy as np

from detime import DecompositionConfig, decompose

t = np.arange(96, dtype=float)
panel = np.column_stack(
    [
        0.03 * t + np.sin(2.0 * np.pi * t / 12.0),
        -0.01 * t + 0.6 * np.sin(2.0 * np.pi * t / 12.0 + 0.4),
    ]
)

result = decompose(
    panel,
    DecompositionConfig(
        method="MSSA",
        params={"window": 24, "rank": 8, "primary_period": 12},
        channel_names=["x0", "x1"],
    ),
)

print(result.components["modes"].shape)

Expected output:

(8, 96, 2)

Next steps

• Use Methods & Chooser to decide whether to stay on the flagship path or move to a wrapper.
• Use detime recommend --length ... --channels ... when you want a machine-readable shortlist.
• Use detime schema --name config when you want the packaged config schema.
• Use Tutorials for step-by-step workflows.
• Use Migration from tsdecomp if you are updating older code.