Demo of the PlateauFinder transformer

What does the PlateauFinder do?

• It searches for time series segments of a given minimum length with a constant given value (i.e. plateaus) and returns their starting points (on the time series index) and lengths,

• The value to search for can also be set to np.nan or np.inf to find missing values,

• The minimum length of segments to consider can also be specified; if set to 1, returns as starting points all locations of the given value.

import numpy as np
import pandas as pd

from sktime.transformations.panel.summarize import PlateauFinder

# generate toy data
X = pd.DataFrame(
    pd.Series(
        [
            pd.Series([np.nan, np.nan, 3, 3, np.nan, 2, 2, 3]),
            pd.Series([0, np.nan, np.nan, np.nan, np.nan, np.nan, 2, np.nan]),
            pd.Series([2, np.nan, np.nan, np.nan, 2, np.nan, 3, 1]),
            pd.Series([1, np.nan, np.nan, 3, np.nan, np.nan, 2, 0]),
        ]
    )
)
X.head()

	0
0	0 NaN 1 NaN 2 3.0 3 3.0 4 NaN 5...
1	0 0.0 1 NaN 2 NaN 3 NaN 4 NaN 5...
2	0 2.0 1 NaN 2 NaN 3 NaN 4 2.0 5...
3	0 1.0 1 NaN 2 NaN 3 3.0 4 NaN 5...

#  find plateaus
t = PlateauFinder()
Xt = t.fit_transform(X)
Xt

	0_nan_starts	0_nan_lengths
0	[0]	[2]
1	[1]	[5]
2	[1]	[3]
3	[1, 4]	[2, 2]

---

Generated using nbsphinx. The Jupyter notebook can be found here.