Source code for LOTUS_regression.predictors.seasonal

import numpy as np



[docs]def add_seasonal_components(basis_df, num_components):
    for column in basis_df:
        n_harmonic = num_components.get(column, 0)

        for i in range(n_harmonic):
            basis_df[column + '_sin' + str(i)] = basis_df[column] * np.sin(2*np.pi * (basis_df.index.dayofyear-1) / 365.25 * (i+1))
            basis_df[column + '_cos' + str(i)] = basis_df[column] * np.cos(2*np.pi * (basis_df.index.dayofyear-1) / 365.25 * (i+1))

    return basis_df