R：查找数据框中的所有重叠答案

【问题标题】：R: Find all overlaps in data frameR：查找数据框中的所有重叠
【发布时间】：2017-03-10 16:29:06
【问题描述】：

我有一个包含重复数据的数据框（示例如下）。第一列是从中获取样本数据的数据。我想创建一个表格来计算日期之间每个可能的交叉点的重叠数。也就是说，以 2014.01.14 为例，我想看看 2014.07.14 样本、2014.04.14 样本等中有多少从那个日期开始的元素。

我一直在研究 setdiff 和 calculate.overlap 函数（来自 dplyr 和维恩图函数），但还没有找到一种有效的方法来进行这些多重集合重叠。

任何帮助将不胜感激 - 谢谢。

编辑：所需的输出是一个汇总所有重叠组合的表格 - 所以一个 x x x 网格，其中 x 是不同日期的数量，元素是每个组合中的重叠数量。换句话说，该表的对角线将全部为 100%，因为它将与自己比较集合。

structure(list(.id = c("2016.10.14", "2016.07.14", "2014.04.14", 
"2014.01.14", "2017.01.06", "2016.10.14", "2016.07.14", "2016.04.14", 
"2016.07.14", "2016.04.14", "2016.01.14", "2015.10.14", "2015.07.14", 
"2015.04.14", "2015.01.14", "2014.10.14", "2017.03.06", "2017.01.06", 
"2016.10.14", "2016.07.14", "2016.04.14", "2016.01.14", "2015.04.14", 
"2015.01.14", "2014.10.14", "2014.07.14", "2014.04.14", "2014.01.14", 
"2016.07.14", "2016.04.14", "2016.01.14", "2015.10.14", "2015.07.14", 
"2015.04.14", "2015.04.14", "2016.01.14", "2015.10.14", "2015.07.14", 
"2015.04.14", "2015.01.14", "2014.10.14", "2014.07.14", "2014.04.14", 
"2014.01.14", "2017.03.06", "2017.01.06", "2016.10.14", "2016.07.14", 
"2015.01.14", "2014.10.14", "2014.07.14", "2014.04.14", "2014.01.14", 
"2016.04.14", "2016.01.14", "2017.03.06", "2017.01.06", "2016.10.14", 
"2016.07.14", "2016.04.14", "2016.01.14", "2015.10.14", "2015.07.14", 
"2015.04.14", "2015.01.14", "2014.10.14", "2014.07.14", "2014.04.14", 
"2014.01.14", "2017.03.06", "2017.01.06", "2015.10.14", "2015.07.14", 
"2017.03.06", "2017.01.06", "2015.10.14", "2015.07.14", "2015.04.14", 
"2015.01.14", "2014.10.14", "2014.07.14", "2014.04.14", "2014.01.14", 
"2016.01.14", "2015.10.14", "2015.07.14", "2015.04.14", "2015.01.14", 
"2014.10.14", "2015.07.14", "2017.03.06", "2017.01.06", "2016.10.14", 
"2016.07.14", "2016.04.14", "2016.01.14", "2015.10.14", "2015.07.14", 
"2015.04.14", "2015.01.14"), Name = c("ZIMMER BIOMET HOLDINGS INC", 
"ZIMMER BIOMET HOLDINGS INC", "ZIMMER BIOMET HOLDINGS INC", "ZIMMER BIOMET HOLDINGS INC", 
"XYLEM INC", "XYLEM INC", "XYLEM INC", "XYLEM INC", "XL GROUP LTD", 
"XL GROUP LTD", "XL GROUP LTD", "XL GROUP LTD", "XL GROUP LTD", 
"XL GROUP LTD", "XL GROUP LTD", "XL GROUP LTD", "XCEL ENERGY INC", 
"XCEL ENERGY INC", "XCEL ENERGY INC", "XCEL ENERGY INC", "XCEL ENERGY INC", 
"XCEL ENERGY INC", "XCEL ENERGY INC", "XCEL ENERGY INC", "XCEL ENERGY INC", 
"XCEL ENERGY INC", "XCEL ENERGY INC", "XCEL ENERGY INC", "WILLIS TOWERS WATSON PLC", 
"WILLIS TOWERS WATSON PLC", "WEYERHAEUSER CO", "WEYERHAEUSER CO", 
"WEYERHAEUSER CO", "WEYERHAEUSER CO", "WELLTOWER INC", "WELLS FARGO & CO", 
"WELLS FARGO & CO", "WELLS FARGO & CO", "WELLS FARGO & CO", "WELLS FARGO & CO", 
"WELLS FARGO & CO", "WELLS FARGO & CO", "WELLS FARGO & CO", "WELLS FARGO & CO", 
"WEC ENERGY GROUP INC", "WEC ENERGY GROUP INC", "WEC ENERGY GROUP INC", 
"WEC ENERGY GROUP INC", "WEC ENERGY GROUP INC", "WEC ENERGY GROUP INC", 
"WEC ENERGY GROUP INC", "WEC ENERGY GROUP INC", "WEC ENERGY GROUP INC", 
"WATERS CORP", "WATERS CORP", "WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", 
"WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", 
"WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", 
"WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", 
"WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", "WASTE MANAGEMENT INC", 
"WALT DISNEY CO/THE", "WALT DISNEY CO/THE", "WALT DISNEY CO/THE", 
"WALT DISNEY CO/THE", "WAL-MART STORES INC", "WAL-MART STORES INC", 
"WAL-MART STORES INC", "WAL-MART STORES INC", "WAL-MART STORES INC", 
"WAL-MART STORES INC", "WAL-MART STORES INC", "WAL-MART STORES INC", 
"WAL-MART STORES INC", "WAL-MART STORES INC", "VORNADO REALTY TRUST", 
"VORNADO REALTY TRUST", "VORNADO REALTY TRUST", "VORNADO REALTY TRUST", 
"VORNADO REALTY TRUST", "VORNADO REALTY TRUST", "VF CORP", "VERIZON COMMUNICATIONS INC", 
"VERIZON COMMUNICATIONS INC", "VERIZON COMMUNICATIONS INC", "VERIZON COMMUNICATIONS INC", 
"VERIZON COMMUNICATIONS INC", "VERIZON COMMUNICATIONS INC", "VERIZON COMMUNICATIONS INC", 
"VERIZON COMMUNICATIONS INC", "VERIZON COMMUNICATIONS INC", "VERIZON COMMUNICATIONS INC"
), X..Wgt..P. = c(0.919761358, 0.933057358, 0.864039801, 0.986472727, 
0.818192145, 0.964415732, 0.937955303, 0.975205821, 0.842865506, 
1.126873419, 1.112947401, 1.099063977, 1.164558843, 1.151605792, 
0.988825771, 0.918127246, 0.961678801, 1.033232496, 1.071266028, 
1.058801025, 0.93020876, 1.006470014, 0.964550593, 1.125389358, 
1.124178769, 1.059881818, 1.085759766, 1.0200896, 0.909300947, 
1.012656226, 0.871314497, 0.94065386, 0.936539398, 0.89578771, 
0.970174408, 0.872847521, 0.949033595, 1.002482323, 0.969762291, 
0.942347801, 0.989303579, 0.98593113, 0.968543074, 0.950396767, 
0.92842183, 0.972461108, 1.006570516, 1.005416119, 0.993368064, 
1.002700203, 0.958855231, 1.096422022, 1.021761901, 0.963388913, 
0.926948746, 1.304745321, 1.199632088, 1.234399398, 1.269585556, 
1.087885176, 1.189337133, 1.146522298, 1.053416881, 1.16547938, 
1.096877887, 1.018844619, 1.014660123, 0.999372549, 1.032241197, 
1.063327494, 1.00757917, 0.963540352, 0.944914081, 0.896543743, 
0.80520587, 0.841142125, 0.921557168, 0.959306858, 1.278648475, 
1.299562929, 1.191759169, 1.222574274, 1.082399965, 0.943414727, 
0.998604251, 0.94817809, 1.097301363, 1.220343667, 0.931967545, 
0.900281393, 1.018445806, 1.18929386, 1.085480841, 1.13721209, 
1.070562566, 1.135912005, 1.0973153, 1.025256697, 1.008197447, 
0.907933413), Mkt.Val..P. = c(60495554.15, 74208256.58, 33088060.26, 
36688942.4, 52333970.68, 63432610.68, 74597801.67, 66667061.11, 
67035085.44, 77035367.75, 61854260.94, 52447251.15, 55756621.76, 
59977265.12, 52416987.1, 41839037.8, 62294332.68, 66088582.56, 
70460485.65, 84208947.55, 63590970.16, 55936568.82, 50235164.73, 
59656130.73, 51228801, 43818397.08, 41578738.06, 37939222.8, 
72318853.08, 69227247.24, 48425032.67, 44887932.18, 44839531.56, 
46653896.1, 50528060.96, 48510233.52, 45287812.5, 47996739.75, 
50506597.32, 49953221.25, 45082541.65, 40761074.48, 37089971.51, 
35347203.52, 60140057.4, 62201466.24, 66205261.4, 79963119.81, 
52657770.96, 45693025.53, 39641683.2, 41987044.93, 38001419.1, 
65859233.16, 51517016.5, 84517033.09, 76731988.76, 81190272.8, 
100973139.4, 74370051.84, 66099771.92, 54711958.7, 50435379, 
60699821.28, 58144756.89, 46428726.18, 41948809.2, 38270574, 
38391165.6, 68878794.6, 64447720.56, 45980073.86, 45240493.7, 
58075101.72, 51503330.42, 40139239.59, 44122213.97, 49962063.55, 
67780293.12, 59221053.22, 49270565.45, 46817995.2, 40256672.96, 
52432146.06, 47653320.52, 45396767.6, 57149013.33, 64689594.6, 
42469739.92, 43103574.75, 65971509.17, 76070725.4, 71395437.96, 
90445164.8, 73185842.88, 63130564.42, 52363804.43, 49087128.75, 
52508354.86, 48128937.78), Ticker = c("ZBH US", "ZBH US", "ZBH US", 
"ZBH US", "XYL US", "XYL US", "XYL US", "XYL US", "XL US", "XL US", 
"XL US", "XL US", "XL US", "XL US", "XL US", "XL US", "XEL US", 
"XEL US", "XEL US", "XEL US", "XEL US", "XEL US", "XEL US", "XEL US", 
"XEL US", "XEL US", "XEL US", "XEL US", "WLTW US", "WLTW US", 
"WY US", "WY US", "WY US", "WY US", "HCN US", "WFC US", "WFC US", 
"WFC US", "WFC US", "WFC US", "WFC US", "WFC US", "WFC US", "WFC US", 
"WEC US", "WEC US", "WEC US", "WEC US", "WEC US", "WEC US", "WEC US", 
"WEC US", "WEC US", "WAT US", "WAT US", "WM US", "WM US", "WM US", 
"WM US", "WM US", "WM US", "WM US", "WM US", "WM US", "WM US", 
"WM US", "WM US", "WM US", "WM US", "DIS US", "DIS US", "DIS US", 
"DIS US", "WMT US", "WMT US", "WMT US", "WMT US", "WMT US", "WMT US", 
"WMT US", "WMT US", "WMT US", "WMT US", "VNO US", "VNO US", "VNO US", 
"VNO US", "VNO US", "VNO US", "VFC US", "VZ US", "VZ US", "VZ US", 
"VZ US", "VZ US", "VZ US", "VZ US", "VZ US", "VZ US", "VZ US"
)), .Names = c(".id", "Name", "X..Wgt..P.", "Mkt.Val..P.", "Ticker"
), row.names = c(NA, 100L), class = "data.frame")

【问题讨论】：

我认为你应该首先清理你的列名、日期等。此外，提供所需的输出。您也可以考虑删除Name 列，该列显然不会增加任何问题（大小除外）。另外，请显示与示例对应的所需输出。您对重叠的描述本身不够清楚。
已编辑以上内容 - 希望对您有所帮助。
好的，您可能想查看table 和expand.grid。使用 dplyr，一个非常丑陋的方法是：DF %>% group_by(Ticker) %>% do({ud = unique(.$.id); expand.grid(ud, ud)}) %>% with(table(Var1, Var2))。至于将其转换为重叠部分，我想您可以找出一些扩展。
谢谢！这就是诀窍

标签： r dplyr overlap intersect

【解决方案1】：

虽然弗兰克在a comment 中发布了解决方案，但这个问题没有得到答案。

因此，我将发布两个解决方案，它们使用dplyr/tidyr 和 cross join/dcast() from data.table 实现 Frank 的方法。

在所有三种方法中，.id 列针对每个 Ticker 进行扩展，然后为 .id/.id 的每个唯一组合计算 Ticker 的出现次数，然后从 long 重新调整为最后是宽格式。

`dplyr`/`tidyr`

library(dplyr)
library(tidyr)
DF %>% 
  group_by(Ticker) %>%
  expand(.id, .id) %>% 
  group_by(.id, .id1) %>% 
  count() %>% 
  spread(.id1, n)

# A tibble: 14 x 15
# Groups:   .id [14]
          .id `2014.01.14` `2014.04.14` `2014.07.14` `2014.10.14` `2015.01.14` `2015.04.14` `2015.07.14` `2015.10.14`
 *      <chr>        <int>        <int>        <int>        <int>        <int>        <int>        <int>        <int>
 1 2014.01.14            6            6            5            5            5            4            3            3
 2 2014.04.14            6            6            5            5            5            4            3            3
 3 2014.07.14            5            5            5            5            5            4            3            3
 4 2014.10.14            5            5            5            7            7            6            5            5
 5 2015.01.14            5            5            5            7            8            7            6            6
 6 2015.04.14            4            4            4            6            7            9            7            7
 7 2015.07.14            3            3            3            5            6            7            9            8
 8 2015.10.14            3            3            3            5            6            7            8            8
 9 2016.01.14            3            3            3            5            6            7            6            6
10 2016.04.14            2            2            2            3            4            4            3            3
11 2016.07.14            4            4            3            4            5            4            3            3
12 2016.10.14            4            4            3            3            4            3            2            2
13 2017.01.06            4            4            4            4            5            4            4            4
14 2017.03.06            4            4            4            4            5            4            4            4
# ... with 6 more variables: `2016.01.14` <int>, `2016.04.14` <int>, `2016.07.14` <int>, `2016.10.14` <int>,
#   `2017.01.06` <int>, `2017.03.06` <int>

`data.table`

library(data.table)
setDT(DF)[, CJ(.id, .id), by = Ticker][, dcast(.SD, V1 ~ V2, length, value.var = "Ticker")]

            V1 2014.01.14 2014.04.14 2014.07.14 2014.10.14 2015.01.14 2015.04.14 2015.07.14 2015.10.14 2016.01.14
 1: 2014.01.14          6          6          5          5          5          4          3          3          3
 2: 2014.04.14          6          6          5          5          5          4          3          3          3
 3: 2014.07.14          5          5          5          5          5          4          3          3          3
 4: 2014.10.14          5          5          5          7          7          6          5          5          5
 5: 2015.01.14          5          5          5          7          8          7          6          6          6
 6: 2015.04.14          4          4          4          6          7          9          7          7          7
 7: 2015.07.14          3          3          3          5          6          7          9          8          6
 8: 2015.10.14          3          3          3          5          6          7          8          8          6
 9: 2016.01.14          3          3          3          5          6          7          6          6          8
10: 2016.04.14          2          2          2          3          4          4          3          3          5
11: 2016.07.14          4          4          3          4          5          4          3          3          4
12: 2016.10.14          4          4          3          3          4          3          2          2          3
13: 2017.01.06          4          4          4          4          5          4          4          4          3
14: 2017.03.06          4          4          4          4          5          4          4          4          3
    2016.04.14 2016.07.14 2016.10.14 2017.01.06 2017.03.06
 1:          2          4          4          4          4
 2:          2          4          4          4          4
 3:          2          3          3          4          4
 4:          3          4          3          4          4
 5:          4          5          4          5          5
 6:          4          4          3          4          4
 7:          3          3          2          4          4
 8:          3          3          2          4          4
 9:          5          4          3          3          3
10:          7          6          4          4          3
11:          6          8          6          5          4
12:          4          6          6          5          4
13:          4          5          5          7          6
14:          3          4          4          6          6

【讨论】：

dplyr/tidyr

data.table

`dplyr`/`tidyr`

`data.table`