Python循环计算时间差

Question

我有以下数据

data = {'timestamp': ['Friday, October 15, 2021 3:40 PM', 'Oct 15, 2021 03:06:29 PM', 'Friday, October 15, 2021 2:28 PM', 'Oct 15, 2021 06:23:51 AM', 'Oct 15, 2021 04:19:07 AM', 'Oct 15, 2021 08:19:07 AM'],
'emailuser': ['[e-mail]', '[e-mail]', '[e-mail]', '[e-mail]','[e-mail]', '[e-mail]']
}
data = pd.DataFrame(data)
print(data)

我想计算谷歌员工的平均响应时间。在本例中，我想要得到

和 之间的时间差

1. [e-mail] - [e-mail](可以跳过[e-mail]的时间戳，因为caron和luke在同一家公司)
2. [e-mail] - [e-mail]要忽略，因为它会导致负时差

Answer 1

看起来不太漂亮，但这应该是符合规格的。如果您有任何问题，或者如果您看到与规格有任何偏差，请评论。

import datetime
import pandas as pd
data = {
"timestamp": [
"Friday, October 15, 2021 3:40 PM",
"Oct 15, 2021 03:06:29 PM",
"Friday, October 15, 2021 2:28 PM",
"Oct 15, 2021 06:23:51 AM",
"Oct 15, 2021 04:19:07 AM",
"Oct 15, 2021 08:19:07 AM",
],
"emailuser": [
"[e-mail]",
"[e-mail]",
"[e-mail]",
"[e-mail]",
"[e-mail]",
"[e-mail]",
],
}

def extract_datetime(timestamp: str) -> datetime.datetime:
for format in ["%A, %B %d, %Y %I:%M %p", "%b %d, %Y %I:%M:%S %p"]:
try:
return datetime.datetime.strptime(timestamp, format)
except:
pass
raise ValueError(f"Timestamp {timestamp} is invalid")

data = pd.DataFrame(data)
data["timestamp"] = data["timestamp"].apply(extract_datetime)
data["delta"] = datetime.timedelta(seconds=0)
gem = "@google.com"
for i in data.index:
if data["emailuser"].iat[i][-len(gem) :] == gem:
for j in range(i+1, data.index[-1]):
if data["emailuser"].iat[j][-len(gem) :] != gem:
delta = data["timestamp"].iat[i] - data["timestamp"].iat[j]
if delta > datetime.timedelta(0):
data["delta"].iat[i] = delta
else:
break

print(data)

timestamp           emailuser           delta
0 2021-10-15 15:40:00  [e-mail] 0 days 01:12:00
1 2021-10-15 15:06:29        [e-mail] 0 days 00:00:00
2 2021-10-15 14:28:00         [e-mail] 0 days 00:00:00
3 2021-10-15 06:23:51      [e-mail] 0 days 00:00:00
4 2021-10-15 04:19:07      [e-mail] 0 days 00:00:00
5 2021-10-15 08:19:07         [e-mail] 0 days 00:00:00