使用tsfresh报错:Can only compute partial correlation for lugs up to 50% of the sample size.

关于很多博客中的使用tsfresh提取时间序列特征,运行例子的时候报错:Can only compute partial correlation for lugs up to 50% of the sample size.

解决方法:

1. 网上博客例子是这样的:

  import tsfresh
  from tsfresh.examples.robot_execution_failures import download_robot_execution_failures, load_robot_execution_failures
  download_robot_execution_failures()
  timeseries, y = load_robot_execution_failures()
  from tsfresh import extract_features
  extraced_features = extract_features(timeseries, column_id='id', column_sort='time')
  extraced_features.head()
替换成这样:

import tsfresh
from tsfresh.examples.robot_execution_failures import download_robot_execution_failures, load_robot_execution_failures

from tsfresh import extract_features
robot_execution_failures.download_robot_execution_failures()
df, y = robot_execution_failures.load_robot_execution_failures()
df.head()

extraction_settings = ComprehensiveFCParameters()

X = extract_features(df, column_id='id', column_sort='time',
                     default_fc_parameters=extraction_settings,
                     # we impute = remove all NaN features automatically
                     impute_function=impute)

X.head()

2. 如果还有错误,先重启你的notebook然后再试试。

3. 如果还有错误,那么首先重新安装你的pandas和numpy,然后重启你的notebook。

4.如果还有错误,手动删掉你安装的tsfresh包,重新安装tsfresh,然后重启你的notebook测试。

经过上面的步骤后,经过我的亲测,可以解决各种报错问题。最后如果成功会出现类似下图效果:

使用tsfresh报错:Can only compute partial correlation for lugs up to 50% of the sample size._第1张图片

有问题可加qq群Python数据分析&机器学习636866908或者群R语言&大数据分析456726635与我进行沟通讨论。

你可能感兴趣的:(python,算法模型,时间序列,tsfresh,特征提取)