清理数据表

Question

我有一个 data.table head(LocalCodes, n= 20) Local Codes 1: Crane, Indiana 0189 2: Rutland, Vermont 0401 3: NA 5003 4: Naval Air Station Patuxent River, Maryland 5001 5: Williamsburg, Virginia 7408 6: District of Columbia, District of Columbia 0132 7: Newport, Rhode Island 1702 8: NA 1805 9: NA 5306 10: Washington DC, District of Columbia / Kansas City, Missouri 2210 11: Kansas City, Missouri 0503 12: Arlington, Virginia 0501 13: Phoenix, Arizona 0301 14: Washington DC, District of Columbia 0132 15: NA 5001 16: Collbran, Colorado 0303 17: Washington DC, District of Columbia / Norfolk, Virginia 1102 18: Minot, North Dakota 1802 19: Washington DC, District of Columbia 2005 20: Pine Knot, Kentucky 4749

我正在尝试使用Good <- LocalCodes[ , list( LocalCodes$Local <- unlist( strsplit( LocalCodes$Local , " / " ) ) , by=LocalCodes$Codes)] 在“/”上拆分Local，并在新数据表中保持相同的Codes。

我不断收到错误Error in strsplit(LocalCodes$Local, " / ") : non-character argument

我确实尝试将as.character(LocalCodes$Local) 添加到Good 以消除错误，但随后 data.table 工作不正确。它将Local 分开，但随后Codes 不排队，因为Local 现在是一个字符。

有没有办法将Local 分开并在正确的Local 上维护Codes

示例： Local Codes 8: NA 1805 9: NA 5306 10: Kansas City, Missouri 2210 11: Washington DC, District of Columbia 2210 12: Kansas City, Missouri 0503 13: Arlington, Virginia 0501 14: Phoenix, Arizona 0301 15: Washington DC, District of Columbia 0132 16: NA 5001 17: Collbran, Colorado 0303 18: Norfolk, Virginia 1102 19: Washington DC, District of Columbia 1102 使用：Plyr、Dplyr、Data.Table

编辑： 这是 dput 输出：

dput(head(LocalCodes, n= 20)) structure(list(Local = list("Crane, Indiana", "Rutland, Vermont", "NA", "Naval Air Station Patuxent River, Maryland", "Williamsburg, Virginia", "District of Columbia, District of Columbia", "Newport, Rhode Island", "NA", "NA", "Washington DC, District of Columbia / Kansas City, Missouri", "Kansas City, Missouri", "Arlington, Virginia", "Phoenix, Arizona", "Washington DC, District of Columbia", "NA", "Collbran, Colorado", "Washington DC, District of Columbia / Norfolk, Virginia", "Minot, North Dakota", "Washington DC, District of Columbia", "Pine Knot, Kentucky"), Codes = list("0189", "0401", "5003", "5001", "7408", "0132", "1702", "1805", "5306", "2210", "0503", "0501", "0301", "0132", "5001", "0303", "1102", "1802", "2005", "4749")), class = c("data.table", "data.frame"), row.names = c(NA, -20L)

Answer 1

我最初的答案没有成功，不止一项包含“/”。我的策略是处理您的 data.table 对象的变体，但在此过程中发现您的结构很遗憾是非标准的。请注意，dput 输出以

开头

structure(list(Local = list("Crane, Indiana",

典型的 data.table 不是列表列表。这种结构因弄乱 data.frame 操作而臭名昭著，而且显然也弄乱了 data.table 操作。这将修复您的数据对象，使其看起来像一个“普通”数据表。

LocalCodes[ , names(LocalCodes) := lapply(LocalCodes,unlist)]
#> dput(LocalCodes)
# structure(list(Local = c("Crane, Indiana", ...

现在它不是列表列表。因此，现在尝试将字符串内部有斜线的情况与没有的情况分开处理，然后将它们 rbind 在一起：

 rbind( LocalCodes[grepl("/",Local) ,
            cbind( data.table(Local=unlist( strsplit(Local, split="/")),
                                     Codes=rep(Codes,each=2)))],
        LocalCodes[!grepl("/",Local)] )
                                         Local Codes
 1:       Washington DC, District of Columbia   2210
 2:                      Kansas City, Missouri  2210
 3:       Washington DC, District of Columbia   1102
 4:                          Norfolk, Virginia  1102
 5:                             Crane, Indiana  0189
 6:                           Rutland, Vermont  0401
 7:                                         NA  5003
 8: Naval Air Station Patuxent River, Maryland  5001
 9:                     Williamsburg, Virginia  7408
10: District of Columbia, District of Columbia  0132
11:                      Newport, Rhode Island  1702
snipped-----