提取特征概述 tsfresh.feature_extraction.feature_calculators 此模块包含特征计算器,它将时间序列作为输入并计算特征值。 以下详细列表包含当前版本tsfresh中计算的所有特征: abs_energy(x) absolute_sum_of_changes(x) agg_autocorrelation(x, param) agg_linear_trend(x, param) approximate_entropy(x, m, r) ar_coefficient(x, param) augmented_dickey_fuller(x, param) autocorrelation(x, lag) binned_entropy(x, max_bins) c3(x, lag) change_quantiles(x, ql, qh, isabs, f_agg) cid_ce(x, normalize) count_above_mean(x) count_below_mean(x) cwt_coefficients(x, param) energy_ratio_by_chunks(x, param) fft_aggregated(x, param) fft_coefficient(x, param) first_location_of_maximum(x) first_location_of_minimum(x) friedrich_coefficients(x, param) has_duplicate(x) has_duplicate_max(x) has_duplicate_min(x) index_mass_quantile(x,param) kurtosis(x) large_standard_deviation(x,r) 返回时间序列的绝对能量,即平方值的和 返回序列x中连续更改的绝对值的和 计算聚合函数 本文来源:https://www.wddqw.com/doc/6bf51ee0a4e9856a561252d380eb6294dc8822c0.html