日志
随机时间序列按要求分组
生成自ChatGPT：要将一个非等距时间序列分组，同时更改序号。将邻近6小时以内的时间点放在同一组中，你可以使用Python来实现。以下是一个示例代码，假设你有一个时间序列的列表，其中每个元素是一个datetime对象：
from datetime import datetime, timedelta
# 你的时间序列
data = ["2019-05-26 12:30:00", "2019-05-26 00:30:00", "2019-05-18 18:30:00", "2019-04-27 12:30:00", "2018-06-05 00:30:00", "1980-04-19 12:30:00", "1980-04-19 06:30:00", "1980-04-19 00:30:00", "1980-04-18 18:30:00", "1980-04-18 12:30:00"]
# 将字符串转换为datetime对象，并按时间排序
timestamps = [datetime.strptime(ts, "%Y-%m-%d %H:%M:%S") for ts in data]
timestamps.sort()
# 初始化一个字典用于存储分组和对应的序号
groups = {}
group_number = 1
# 定义一个函数来判断两个时间是否在6小时内
def within_six_hours(time1, time2):
return (time2 - time1) <= timedelta(hours=6)
# 遍历时间序列并分组
current_group = []
for timestamp in timestamps:
if not current_group:
current_group.append(timestamp)
else:
if within_six_hours(current_group[-1], timestamp):
current_group.append(timestamp)
else:
group_key = current_group[0]
group_value = current_group
groups[group_key] = (group_number, group_value)
current_group = [timestamp]
group_number += 1
# 处理最后一个分组
if current_group:
group_key = current_group[0]
group_value = current_group
groups[group_key] = (group_number, group_value)
# 打印分组结果
for key, (group_number, value) in groups.items():
print(f'Group {group_number} starting at {key}:')
for timestamp in value:
print(f' {timestamp}')
import pandas as pd
# 之前生成的字典
groups = {
datetime(2018, 6, 5, 0, 30): (1, [datetime(2018, 6, 5, 0, 30)]),
datetime(1980, 4, 18, 12, 30): (2, [datetime(1980, 4, 18, 12, 30)]),
datetime(1980, 4, 18, 18, 30): (2, [datetime(1980, 4, 18, 18, 30)]),
# ... 其他分组数据 ...
}
# 创建一个空的DataFrame
df = pd.DataFrame(columns=['Group', 'Timestamp'])
# 遍历字典并将数据添加到DataFrame
for key, (group_number, value) in groups.items():
for timestamp in value:
df = df.append({'Group': group_number, 'Timestamp': timestamp}, ignore_index=True)
# 打印DataFrame
print(df)