[英]Create a list from a list of dictionaries subject to a condition (counter and accumulator) in python
[英]Create a new column from loop condition (while, list id, counter, timestamp...)
我想用这个数据框创建一个新的时间戳列:
Timestamp Flag
0 2019-10-21 07:48:28.272688 end
1 2019-10-21 07:48:28.449916 end
2 2019-10-21 07:48:26.740378 begin
3 2019-10-21 07:48:26.923764 begin
4 2019-10-21 07:48:41.689466 end
5 2019-10-21 07:48:37.306045 begin
6 2019-10-21 07:58:00.774449 end
7 2019-10-21 07:57:59.223986 begin
8 2019-10-21 08:32:37.004455 end
9 2019-10-21 08:32:35.755252 begin
原理很简单:
对于每一行,如果我有一个结束 => 计数器 +=1 否则(我有一个开始)=> 计数器 -=1
当 counter == 0 => 将时间戳的 id 保存在列表中
所以结果一定是:
Timestamp Flag
0 2019-10-21 07:48:28.272688 end
2 2019-10-21 07:48:26.740378 begin
3 2019-10-21 07:48:26.923764 begin
4 2019-10-21 07:48:41.689466 end
5 2019-10-21 07:48:37.306045 begin
6 2019-10-21 07:58:00.774449 end
7 2019-10-21 07:57:59.223986 begin
8 2019-10-21 08:32:37.004455 end
9 2019-10-21 08:32:35.755252 begin
因为: First end => counter = 1 (save(first row), ct = 2, ct = 1(save), ct = 0 (save), (save) ct = 1; ct =0 (save)...
目前我无法将相应的值添加到 ID 中,也许我在代码中忘记了 (a) 个条件。
我的一段代码:
counter = 0
i = 0
while i < len(df):
id_timestamp_to_save = []
if df.loc[i, 'Flag'] == 'end':
counter +=1
if counter == 1:
id_timestamp_to_save = list(range(i))
else:
counter -=1
if counter == 0:
id_timestamp_to_save = list(range(i))
df['New_Timestamp'] = df['New_Timestamp'].assign(id_timestamp_to_save)
i+=1
请帮帮我。
根据您的逻辑,只需将end
替换为1
并以-1
begin
,然后cumsum
:
counters = df.Flag.map({'end':1,'begin':-1}).cumsum().eq(0)
df[counters | counters.shift(fill_value=True)]
输出:
Timestamp Flag
0 2019-10-21 07:48:28.272688 end
3 2019-10-21 07:48:26.923764 begin
4 2019-10-21 07:48:41.689466 end
5 2019-10-21 07:48:37.306045 begin
6 2019-10-21 07:58:00.774449 end
7 2019-10-21 07:57:59.223986 begin
8 2019-10-21 08:32:37.004455 end
9 2019-10-21 08:32:35.755252 begin
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.