Question

我正在尝试使用包含字符串的列合并两个 data.frames。两列中的字符串是名称，不幸的是，它们的顺序不同。在下面的示例中，df_1 中的名称具有“name”+“midname”+“surname1”+“surname2”结构，而在 df_2 中，结构为“surname1”+“surname2”+“name “+”中间名。

我首先尝试使用名称执行 fuzzy merge。但是，它并没有解决问题，因为在完全不同的名称之间仍然存在非零匹配。此外，定义一个可以定义名称何时与另一个名称完全不同的切割点并非易事。我还期待具有相反顺序的名称之间的相似度更高（即，（姓名+中间名）+（姓氏1+姓氏2）以不同的顺序）。

您有更好的方法以不同的顺序使用这些名称来合并两个 data.frame 吗？提前致谢。

# "name"+"midname"+"surname1"+"surname2
df_1<- read.table(header = T,sep = "\t", text = "
name
Tetsurō Shoyo Hinata Kuroo
Kōtarō Tobio Kageyama Bokuto
Wakatoshi Daichi Sawamura Ushijima
Tōru  Tsukishima Oikawa
Yūji  Azumane Terushima
Kenma Kozume
")
# "surname1"+"surname2"+"name"+"midname".
df_2<- read.table(header = T,sep = "\t", text = "
name
Hinata Kuroo Tetsurō Shoyo
Kageyama Bokuto Kōtarō Tobio
Sawamura Ushijima Wakatoshi Daichi
Tsukishima Oikawa Tōru 
Azumane Terushima Yūji 
Kiyoomi Sakusa
")
library(fuzzyjoin); library(dplyr);
stringdist_join(df_1, df_2, 
                by = "name",
                mode = "inner",
                ignore_case = FALSE, 
                method = "jw", 
                max_dist = 99, 
                distance_col = "dist") %>%
  group_by(name.x) %>%
  slice_min(order_by = dist, n = 1)

结果

# A tibble: 6 x 3
# Groups:   name.x [6]
name.x                          name.y                           dist
<chr>                           <chr>                           <dbl>
1 Kenma Kozume                    "Azumane Terushima Yuji "       0.416
2 Kotaro Tobio Kageyama Bokuto    "Kageyama Bokuto Kotaro Tobio"  0.241
3 Tetsuro Shoyo Hinata Kuroo      "Kageyama Bokuto Kotaro Tobio"  0.351
4 Toru  Tsukishima Oikawa         "Tsukishima Oikawa Toru "       0.302
5 Wakatoshi Daichi Sawamura Ushi~ "Sawamura Ushijima Wakatoshi D~ 0.366
6 Yuji  Azumane Terushima         "Azumane Terushima Yuji "       0.283

Answer 1

您可以strsplit到个人名称，sort他们和paste。然后使用 match。

x <- sapply(strsplit(df_1$name, " +"), function(x) paste(sort(x), collapse = " "))
y <- sapply(strsplit(df_2$name, " +"), function(x) paste(sort(x), collapse = " "))

cbind(df_1$name, df_2$name[match(x, y)])
#     [,1]                                 [,2]                                
#[1,] "Tetsurō Shoyo Hinata Kuroo"         "Hinata Kuroo Tetsurō Shoyo"        
#[2,] "Kōtarō Tobio Kageyama Bokuto"       "Kageyama Bokuto Kōtarō Tobio"      
#[3,] "Wakatoshi Daichi Sawamura Ushijima" "Sawamura Ushijima Wakatoshi Daichi"
#[4,] "Tōru  Tsukishima Oikawa"            "Tsukishima Oikawa Tōru "           
#[5,] "Yūji  Azumane Terushima"            "Azumane Terushima Yūji "           
#[6,] "Kenma Kozume"                       NA

使用具有相同字符串但顺序不同的列合并两个 data.frame

结果

1 个答案: