Box-Cox Transformation¶

In [ ]:

from typing import Any, Dict

import matplotlib.pyplot as plt
import pandas as pd
from darts import TimeSeries
from darts.dataprocessing.transformers import BoxCox
from darts.datasets import AirPassengersDataset
from darts.utils import statistics as dus

from typing import Any, Dict import matplotlib.pyplot as plt import pandas as pd from darts import TimeSeries from darts.dataprocessing.transformers import BoxCox from darts.datasets import AirPassengersDataset from darts.utils import statistics as dus

In [ ]:

ap_series = AirPassengersDataset().load()

_, ax = plt.subplots(figsize=(10, 6.18))
ap_series.plot(label=f"Air Passenger Original Data", ax=ax)

ap_series = AirPassengersDataset().load() _, ax = plt.subplots(figsize=(10, 6.18)) ap_series.plot(label=f"Air Passenger Original Data", ax=ax)

In [ ]:

_, ax = plt.subplots(figsize=(10, 6.18))

boxcox_opt = BoxCox()
ap_boxcox_opt_transformed = boxcox_opt.fit_transform(ap_series)
ap_boxcox_opt_transformed.plot(
    label=f"$\lambda={boxcox_opt._fitted_params[0].item():0.3f}$", ax=ax
)

_, ax = plt.subplots(figsize=(10, 6.18)) boxcox_opt = BoxCox() ap_boxcox_opt_transformed = boxcox_opt.fit_transform(ap_series) ap_boxcox_opt_transformed.plot( label=f"$\lambda={boxcox_opt._fitted_params[0].item():0.3f}$", ax=ax )

In [ ]:

_, ax = plt.subplots(figsize=(10, 6.18))
lmbda = 0.01
boxcox = BoxCox(lmbda=lmbda)
boxcox_transformed = boxcox.fit_transform(ap_series)
boxcox_transformed.plot(label=f"Box-Cox Transformed Data (lambdax={lmbda})", ax=ax)

_, ax = plt.subplots(figsize=(10, 6.18)) lmbda = 0.01 boxcox = BoxCox(lmbda=lmbda) boxcox_transformed = boxcox.fit_transform(ap_series) boxcox_transformed.plot(label=f"Box-Cox Transformed Data (lambdax={lmbda})", ax=ax)

In [ ]:

_, ax = plt.subplots(figsize=(10, 6.18))

for lmbda in [0.01, 0.1, 0.2]:
    boxcox_lmbda = BoxCox(lmbda=lmbda)
    boxcox_lmbda_transformed = boxcox_lmbda.fit_transform(ap_series)
    boxcox_lmbda_transformed.plot(
        label=f"Box-Cox Transformed Data (lambda={lmbda})", ax=ax
    )

_, ax = plt.subplots(figsize=(10, 6.18)) for lmbda in [0.01, 0.1, 0.2]: boxcox_lmbda = BoxCox(lmbda=lmbda) boxcox_lmbda_transformed = boxcox_lmbda.fit_transform(ap_series) boxcox_lmbda_transformed.plot( label=f"Box-Cox Transformed Data (lambda={lmbda})", ax=ax )

Plot and Check Variance¶

In [ ]:

def var_series(series: TimeSeries, window: int = 36) -> TimeSeries:
    series_rolling_var = TimeSeries.from_dataframe(
        series.pd_dataframe().rolling(window=window).var().dropna()
    )

    return series_rolling_var


def check_stationary(series: TimeSeries) -> Dict[str, Any]:
    return {
        "is_stationary": dus.stationarity_tests(series),
        "kpss": dus.stationarity_test_kpss(series),
        "adf": dus.stationarity_test_adf(series),
    }

def var_series(series: TimeSeries, window: int = 36) -> TimeSeries: series_rolling_var = TimeSeries.from_dataframe( series.pd_dataframe().rolling(window=window).var().dropna() ) return series_rolling_var def check_stationary(series: TimeSeries) -> Dict[str, Any]: return { "is_stationary": dus.stationarity_tests(series), "kpss": dus.stationarity_test_kpss(series), "adf": dus.stationarity_test_adf(series), }

In [ ]:

boxcox_grid = []
lmbdas = [0.01, 0.1]
rolling_window = 12

for idx, lmbda in enumerate(lmbdas):
    boxcox_lmbda = BoxCox(lmbda=lmbda)
    boxcox_lmbda_transformed = boxcox_lmbda.fit_transform(ap_series)
    var_lmbda_series = var_series(boxcox_lmbda_transformed, window=rolling_window)
    _, ax = plt.subplots(figsize=(10, 6.18))
    var_lmbda_series.plot(
        label=f"Variance (Rolling Window={rolling_window}) (Box-Cox lambda={lmbda})",
        ax=ax,
    )
    plt.show()

boxcox_grid = [] lmbdas = [0.01, 0.1] rolling_window = 12 for idx, lmbda in enumerate(lmbdas): boxcox_lmbda = BoxCox(lmbda=lmbda) boxcox_lmbda_transformed = boxcox_lmbda.fit_transform(ap_series) var_lmbda_series = var_series(boxcox_lmbda_transformed, window=rolling_window) _, ax = plt.subplots(figsize=(10, 6.18)) var_lmbda_series.plot( label=f"Variance (Rolling Window={rolling_window}) (Box-Cox lambda={lmbda})", ax=ax, ) plt.show()

In [ ]:

_, ax = plt.subplots(figsize=(10, 6.18))
var_ap_series = var_series(ap_series, window=rolling_window)
var_ap_series.plot(label=f"Variance (Rolling Window={rolling_window})", ax=ax)

_, ax = plt.subplots(figsize=(10, 6.18)) var_ap_series = var_series(ap_series, window=rolling_window) var_ap_series.plot(label=f"Variance (Rolling Window={rolling_window})", ax=ax)

---

Contributors: LM