在時間範圍內搜索

Question

我有這個帶時間戳的列表，我希望能夠根據用戶輸入在特定時間範圍內搜索所有元素（每個時間段都有另一個列表中的相應信息）（小時數< = 24或從午夜開始的日子或兩者都不）。

實例（這只是一個例子列表，該解決方案應該工作在非常大的列表）

list = ['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31  19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14'] 
activity= [a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w] 

我將使用最後一個元素list[-1]作爲參考點。如果用戶想要查看過去三小時內的活動，那麼女巫的意思是從2002-03-31 16:54:14 to 2002-03-31 19:54:14開始，時間戳的位置將用於從另一個列表中獲取活動。 我首先想到的是將每個時間戳轉換爲可用的東西，以便更容易地比較每個元素，但必須有一個更簡單的解決方案。

這個module看起來可用，但我無法弄清楚如何使用它。

問候

Answer 1

你是非常幸運，您的時間戳以最簡單的順序進行排序，你可以踢在整個「轉換爲時間值」，只是做字符串比較：

times = ['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31  19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14'] 
activity= ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w'] 

start = '2002-03-31 16:54:14' 
end = '2002-03-31 19:54:14' 

for time, activity in zip(times, activity): 
    if time >= start and time <= end: 
     print time, activity 

Answer 2

這樣的工作流程：

• 使用datetime模塊到你的字符串與strptime方法轉化爲datetime對象：你datetime對象的列表。
• 計算timedeltas通過從最後一個減去此列表的每個條目。
• 您可以使用timedelta的seconds屬性找出一點與參考之間的秒數：與3*3600（3h）比較以確定哪些項目處於適當的時段。

Answer 3

像這樣的東西應該工作

ls = ['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31  19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14'] 

# target is one of the items in the list 
target = datetime.strptime('2002-03-31 19:53:17', '%Y-%m-%d %H:%M:%S') 
for l in ls: 
    print datetime.strptime(l, '%Y-%m-%d %H:%M:%S') - target 

打印

-1 day, 23:37:24 
-1 day, 23:37:24 
-1 day, 23:50:32 
-1 day, 23:50:33 
-1 day, 23:56:48 
-1 day, 23:56:49 
-1 day, 23:56:49 
-1 day, 23:57:27 
-1 day, 23:57:28 
-1 day, 23:57:28 
-1 day, 23:58:33 
-1 day, 23:58:33 
-1 day, 23:58:33 
-1 day, 23:59:08 
-1 day, 23:59:08 
-1 day, 23:59:08 
-1 day, 23:59:48 
-1 day, 23:59:49 
-1 day, 23:59:49 
-1 day, 23:59:49 
0:00:00 
0:00:57 
0:00:57 

的datetime.strptime(l, '%Y-%m-%d %H:%M:%S') - target回報timedelta對象（docs）。您可以訪問timedelta對象days，seconds和microseconds屬性，並將這些屬性與某些所需的時間範圍進行比較。例如，爲了獲得所有發生不到一小時，從一些參考點的事件的所有指標：

less_than_an_hour = [] 
for i,l in enumerate(ls): 
    if (datetime.strptime(l, '%Y-%m-%d %H:%M:%S') - target).seconds < 3600: 
     less_than_an_hour.append(i) 

Answer 4

我倒是：

• 時間戳列表轉換爲datetime個對象：

  times = [datetime.datetime.strptime(t, '%Y-%m-%d %H:%M:%S') for t in times] 
  

• 使用bisect module找到用戶請求的開始時間。使用bisect是一個比使用線性搜索快很多方法，只要你的用戶輸入轉換爲datetime對象，以及：

  start = datetime.datetime(2002, 3, 31, 19, 53, 17) 
  startindex = bisect.bisect_left(times, start) 
  

• 使用itertools功能的兩個列表合併成一個，顯示符合您範圍的條目：

  end = datetime.datetime(2002, 4, 1, 07, 53, 17) 
  
  merged = itertools.izip(times, activity) 
  fromstart = itertools.islice(merged, startindex) 
  untilend = itertools.takewhile(lambda e: e[0] <= end, fromstart) 
  

untilend的迭代現在生成start和end之間條目(time, activity)元組，機智使用任何額外的內存複製列表。這讓您可以高效地處理大量數據。

演示：

>>> import itertools 
>>> import datetime 
>>> import bisect 
>>> times = ['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31  19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14'] 
>>> activity= ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w'] 
>>> times = [datetime.datetime.strptime(t, '%Y-%m-%d %H:%M:%S') for t in times] 
>>> start = datetime.datetime(2002, 3, 31, 19, 53, 17) 
>>> end = datetime.datetime(2002, 4, 1, 07, 53, 17) 
>>> startindex = bisect.bisect_left(times, start) 
>>> merged = itertools.izip(times, activity) 
>>> fromstart = itertools.islice(merged, startindex) 
>>> untilend = itertools.takewhile(lambda e: e[0] <= end, fromstart) 
>>> for time, activity in untilend: 
...  print time, activity 
... 
2002-03-31 19:30:41 a 
2002-03-31 19:30:41 b 
2002-03-31 19:43:49 c 
2002-03-31 19:43:50 d 
2002-03-31 19:50:05 e 
2002-03-31 19:50:06 f 
2002-03-31 19:50:06 g 
2002-03-31 19:50:44 h 
2002-03-31 19:50:45 i 
2002-03-31 19:50:45 j 
2002-03-31 19:51:50 k 
2002-03-31 19:51:50 l 
2002-03-31 19:51:50 m 
2002-03-31 19:52:25 n 
2002-03-31 19:52:25 o 
2002-03-31 19:52:25 p 
2002-03-31 19:53:05 q 
2002-03-31 19:53:06 r 
2002-03-31 19:53:06 s 
2002-03-31 19:53:06 t