删除特定时间段内出现的重复项

Question

我有一个包含 ID 变量、日期和几个代理的数据集(参见下面的示例)。代理已经对每位患者进行了多次测试，我想过滤第一个出现的每个 ID，并删除第一个出现后 4 周内出现的所有其他测试。在此之后，我再次想过滤第一个并删除 4 周内出现的所有其他人 - 在整个数据集中。我还生成了显示周、月和年的变量。

ID <- rep(1, times = 20)
Date <- c("2010-12-09", "2010-12-09", "2010-12-09", "2010-12-09", "2010-12-09", "2010-12-09", "2010-12-09", "2010-12-09", "2010-12-27", "2010-12-27", "2010-12-27", "2010-12-27", "2011-01-14", "2011-01-14", "2011-01-14", "2011-01-14", "2011-01-14", "2011-01-14", "2011-01-14", "2011-01-14")
Agent <- c("Agent1", "Agent2", "Agent3", "Agent4", "Agent1", "Agent2", "Agent3", "Agent4", "Agent1", "Agent2", "Agent3", "Agent4", "Agent1", "Agent2", "Agent3", "Agent4", "Agent1", "Agent2", "Agent3", "Agent4")

df <- data.frame(ID, Date, Agent)


     ID Date        Week Month Year  Agent
1    1  2010-12-09  49   12    2010  Agent1
2    1  2010-12-09  49   12    2010  Agent2 
3    1  2010-12-09  49   12    2010  Agent3 
4    1  2010-12-09  49   12    2010  Agent4 
5    1  2010-12-09  49   12    2010  Agent1 
6    1  2010-12-09  49   12    2010  Agent2 
7    1  2010-12-09  49   12    2010  Agent3
8    1  2010-12-09  49   12    2010  Agent4
9    1  2010-12-27  52   12    2010  Agent1
10   1  2010-12-27  52   12    2010  Agent2
11   1  2010-12-27  52   12    2010  Agent3
12   1  2010-12-27  52   12    2010  Agent4
13   1  2011-01-14  2    1     2011  Agent1
14   1  2011-01-14  2    1     2011  Agent2
15   1  2011-01-14  2    1     2011  Agent3
16   1  2011-01-14  2    1     2011  Agent4
17   1  2011-01-14  2    1     2011  Agent1
18   1  2011-01-14  2    1     2011  Agent2
19   1  2011-01-14  2    1     2011  Agent3
20   1  2011-01-14  2    1     2011  Agent4

我需要的是这个:

     ID Date        Week Month Year  Agent
1    1  2010-12-09  49   12    2010  Agent1
2    1  2010-12-09  49   12    2010  Agent2 
3    1  2010-12-09  49   12    2010  Agent3 
4    1  2010-12-09  49   12    2010  Agent4 
13   1  2011-01-14  2    1     2011  Agent1
14   1  2011-01-14  2    1     2011  Agent2
15   1  2011-01-14  2    1     2011  Agent3
16   1  2011-01-14  2    1     2011  Agent4

我很高兴有任何帮助!

Answer 1

您可以减去最小值 Date每个ID创建一个由 4 周数据组成的新组，并为每个 ID 选择具有最小日期的行, group和 Agent .

library(dplyr)

df %>%
  mutate(Date = as.Date(Date)) %>%
  group_by(ID) %>%
  mutate(group = ceiling(as.integer(difftime(Date, min(Date), units = 'week')/4))) %>%
  group_by(ID, group, Agent) %>%
  slice(which.min(Date))

#     ID Date       Agent  group
#  <dbl> <date>     <chr>  <dbl>
#1     1 2010-12-09 Agent1     0
#2     1 2010-12-09 Agent2     0
#3     1 2010-12-09 Agent3     0
#4     1 2010-12-09 Agent4     0
#5     1 2011-01-14 Agent1     1
#6     1 2011-01-14 Agent2     1
#7     1 2011-01-14 Agent3     1
#8     1 2011-01-14 Agent4     1