去极值

# 缺失数据填充

定义

M.winsorize.v6(columns_input='',median_deviate=)

参数:

  • columns_input:指定去极值列
  • median_deviate:默认值3,最大值100,最小值-2147483648,指定标准差倍数

返回:

  • data,去极值数据

返回类型:Outputs

工作流截图:

示例代码:

#m8为基础特征抽取模块
#m7为输入特征列表模块
m11 = M.winsorize.v6(
    input_data=m8.data,
    features=m7.data,
    columns_input='',
    median_deviate=3
)

运行结果:

[2019-07-17 10:03:17.585114] INFO: bigquant: instruments.v2 开始运行..
[2019-07-17 10:03:17.643444] INFO: bigquant: 命中缓存
[2019-07-17 10:03:17.645984] INFO: bigquant: instruments.v2 运行完成[0.060855s].
[2019-07-17 10:03:17.654485] INFO: bigquant: input_features.v1 开始运行..
[2019-07-17 10:03:17.806171] INFO: bigquant: 命中缓存
[2019-07-17 10:03:17.819613] INFO: bigquant: input_features.v1 运行完成[0.165109s].
[2019-07-17 10:03:17.934718] INFO: bigquant: general_feature_extractor.v7 开始运行..
[2019-07-17 10:03:18.016053] INFO: bigquant: 命中缓存
[2019-07-17 10:03:18.020187] INFO: bigquant: general_feature_extractor.v7 运行完成[0.085451s].
[2019-07-17 10:03:18.024479] INFO: bigquant: derived_feature_extractor.v3 开始运行..
[2019-07-17 10:03:18.131715] INFO: bigquant: 命中缓存
[2019-07-17 10:03:18.134645] INFO: bigquant: derived_feature_extractor.v3 运行完成[0.110148s].
[2019-07-17 10:03:18.138480] INFO: bigquant: standardlize.v8 开始运行..
[2019-07-17 10:03:18.220652] INFO: bigquant: 命中缓存
[2019-07-17 10:03:18.224188] INFO: bigquant: standardlize.v8 运行完成[0.085677s].
[2019-07-17 10:03:18.241237] INFO: bigquant: winsorize.v6 开始运行..
[2019-07-17 10:06:47.225497] INFO: bigquant: winsorize.v6 运行完成[208.98359s].