DeTime Method Gallery

Explore DeTime's decomposition methods. Each card links to a dedicated page with executable code, output, and a decomposition plot. All examples use compact synthetic data and run quickly.

Tip
Regenerate all assets with python scripts/generate_notebook_gallery.py. Jump to downloads for the full notebook, Python source, or zipped example.

Flagship Methods

SSA

→

Univariate SSA

univariate flagship

STD

→

Seasonal-trend decomposition

channelwise flagship

STDR

→

Robust seasonal-trend decomposition

channelwise flagship

MSSA

→

Multivariate SSA

multivariate flagship

Wrapper Methods

STL

→

Statsmodels STL wrapper

univariate

MSTL

→

Statsmodels MSTL wrapper

univariate

ROBUST_STL

→

Robust STL wrapper

univariate

Empirical & Adaptive Methods

EMD

→

Empirical mode decomposition

univariate

CEEMDAN

→

Noise-assisted EMD

univariate

VMD

→

Variational mode decomposition

univariate

WAVELET

→

Wavelet decomposition

univariate

MA_BASELINE

→

Moving-average baseline

univariate

Multivariate & Experimental

MVMD

→

Optional multivariate VMD backend

multivariate

MEMD

→

Optional multivariate EMD backend

multivariate

GABOR_CLUSTER

→

Experimental Gabor clustering path

experimental
## Summary | Method | Status | Input mode | Trend shape | Residual RMSE | |---|---|---|---|---:| | `SSA` | ran | `univariate` | `(96,)` | 0.00000000 | | `STD` | ran | `channelwise` | `(96,)` | 0.00000000 | | `STDR` | ran | `channelwise` | `(96,)` | 0.00000000 | | `MSSA` | ran | `multivariate` | `(96, 3)` | 0.00000000 | | `STL` | ran | `univariate` | `(96,)` | 0.00000000 | | `MSTL` | ran | `univariate` | `(96,)` | 0.00000000 | | `ROBUST_STL` | ran | `univariate` | `(96,)` | 0.00000000 | | `EMD` | ran | `univariate` | `(96,)` | 0.00000000 | | `CEEMDAN` | ran | `univariate` | `(96,)` | 0.00000000 | | `VMD` | ran | `univariate` | `(96,)` | 0.03421948 | | `WAVELET` | ran | `univariate` | `(96,)` | 0.00000000 | | `MA_BASELINE` | ran | `univariate` | `(96,)` | 0.00000000 | | `MVMD` | ran | `multivariate` | `(96, 3)` | 0.08323657 | | `MEMD` | ran | `multivariate` | `(96, 3)` | 0.00000000 | | `GABOR_CLUSTER` | skipped | `univariate` | `` | n/a | Total running time of the gallery script: 4.716 seconds. Methods run: 14; skipped with explicit reason: 1.

Downloads

Download Jupyter notebook: de_time_method_gallery.ipynb Download Python source code: de_time_method_gallery.py Download zipped example: de_time_method_gallery.zip

The GitHub-rendered notebook is also available at examples/notebooks/de_time_method_gallery.ipynb.