获取多个（> 2）字符同时出现的实例数 [英] Get number of instances of co-occurence of multiple (>2) characters

问题描述

这是问题的延伸此处

This is an extension of the question here.

我有一个这样的数据框：

I have a dataframe like so:

df<-structure(list(person = c("p1", "p1", "p1", "p1", "p1", "p1", 
"p1", "p2", "p2", "p2", "p3", "p3", "p3", "p4", "p4", "p4", "p5", 
"p5", "p5", "p6", "p6", "p6", "p7", "p7", "p7"), hp_char = c("hp1", 
"hp2", "hp3", "hp4", "hp5", "hp6", "hp7", "hp8", "hp9", "hp10", 
"hp1", "hp2", "hp3", "hp5", "hp6", "hp7", "hp8", "hp9", "hp10", 
"hp3", "hp4", "hp5", "hp1", "hp2", "hp3")), class = c("tbl_df", 
"tbl", "data.frame"), row.names = c(NA, -25L), .Names = c("person", 
"hp_char"), spec = structure(list(cols = structure(list(person = structure(list(), class = c("collector_character", 
"collector")), hp_char = structure(list(), class = c("collector_character", 
"collector"))), .Names = c("person", "hp_char")), default = structure(list(), class = c("collector_guess", 
"collector"))), .Names = c("cols", "default"), class = "col_spec"))

遵循 Uwe 我得到任何两个 hp_id共现的实例数，如下所示：

Following the really efficient self-join/data.table answer provided by Uwe I get the number of instances of co-ocurrence of any TWO "hp_id" as follows:

df_by2<- setDT(df)[df, on = "person", allow = TRUE][
    hp_char < i.hp_char, .N, by = .(HP_ID1 = hp_char, HP_ID2 = i.hp_char)]

这给我：

    HP_ID1 HP_ID2 N
 1:    hp1    hp2 3
 2:    hp1    hp3 3
 3:    hp2    hp3 3
 4:    hp1    hp4 1
 5:    hp2    hp4 1
 6:    hp3    hp4 2
 7:    hp1    hp5 1
 8:    hp2    hp5 1
 9:    hp3    hp5 2
10:    hp4    hp5 2
11:    hp1    hp6 1
12:    hp2    hp6 1
13:    hp3    hp6 1
14:    hp4    hp6 1
15:    hp5    hp6 2
16:    hp1    hp7 1
17:    hp2    hp7 1
18:    hp3    hp7 1
19:    hp4    hp7 1
20:    hp5    hp7 2
21:    hp6    hp7 2
22:   hp10    hp8 2
23:    hp8    hp9 2
24:   hp10    hp9 2

但是我想知道是否可以扩展此方法，其中
是实例的数量可以计算出两个以上的 hp_char共存。换句话说，我正在寻找这样的输出（例如，发生3次事件的次数）：

However I was wondering if an extension of this method could be made where the number of instance of co-ocurrence of greater than two "hp_char" could be calculated. In other words I was looking for an output (e.g. for number of times 3 events occurring) in like so:

  HP_ID1 HP_ID2 HP_ID3 N
1    hp1    hp2    hp3 3
2    hp3    hp4    hp5 2
3    hp5    hp6    hp7 2
4    hp8    hp9   hp10 2

到目前为止，我已经能够找到两个事件同时发生的多个解决方案，但是它们似乎不能普遍用于计数> 2个事件的实例。谢谢你的帮助！

So far I have been able to find multiple solutions for cooccurrence of two events but they do not seem to be generalizable to counting instances of >2 events. Thanks for any help!

推荐答案

您可以进行 double 自联接，其余部分几乎相同：

You can do a double self-join, the rest of it is pretty much the same:

df2 <- setDT(df)[df, on = "person", allow = TRUE][df, 
                     on = "person", allow = TRUE]

df2[hp_char < i.hp_char & i.hp_char < i.hp_char.1,
    .N,  by = .(HP_ID1 = hp_char, 
                HP_ID2 = i.hp_char,
                HP_ID3 = i.hp_char.1)][N >= 2]
#   HP_ID1 HP_ID2 HP_ID3 N
#1:    hp1    hp2    hp3 3
#2:    hp3    hp4    hp5 2
#3:    hp5    hp6    hp7 2
#4:   hp10    hp8    hp9 2

这篇关于获取多个（> 2）字符同时出现的实例数的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！