R中的混合排序

Question

我在重新调整字母数字列时遇到了一些问题，我正在寻找一些技巧（首选基数 R）。

考虑以下几点：

structure(list(Company = structure(c(2L, 2L, 2L, 2L, 2L, 2L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 3L, 3L, 3L, 3L), .Label = c("ABC Inc", 
"ACME", "Handy Co"), class = "factor"), Quarter = structure(c(2L, 
3L, 4L, 5L, 6L, 1L, 2L, 3L, 2L, 3L, 4L, 5L, 6L, 1L, 3L, 4L, 5L, 
6L), .Label = c("1Q 2013", "1Q 2014", "1Q 2015", "2Q 2014", "3Q 2014", 
"4Q 2014"), class = "factor"), Revenue = c(5000L, 6000L, 3200L, 
1200L, 7200L, 10000L, 2500L, 4100L, 1250L, 8100L, 2300L, 3700L, 
1100L, 1600L, 8000L, 9000L, 10000L, 12000L)), .Names = c("Company", 
"Quarter", "Revenue"), class = "data.frame", row.names = c(NA, 
-18L))

我希望重新调整 Quarter 列以按年份和季度排序（例如 c("1Q 2013", "1Q 2014", "2Q 2014", ..., "1Q 2015")） .假设这只是一个更大的数据集的样本，该数据集涉及可追溯至数十年的季度。

Answer 1

我认为您应该将Quarter 分成两个字段：quarter 和year。这是使用dplyr的解决方案：

library(dplyr)

df %>%
  mutate(
    quarter = as.numeric(substr(Quarter, 1, 1)),
    year = as.numeric(substr(Quarter, 4, 7))
  ) %>%
  select(-Quarter) %>% # original field no longer needed
  arrange(quarter, year)

使用tidyr::separate()可以得到更简洁的解决方案：

library(tidyr)

df %>%
  separate(Quarter, c("quarter", "year")) %>%
  arrange(quarter, year)

Answer 2

这里有两种使用base r的方法

dat <- structure(list(Company = structure(c(2L, 2L, 2L, 2L, 2L, 2L, 
                                     1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 3L, 3L, 3L, 3L), .Label = c("ABC Inc", 
                                                                                                 "ACME", "Handy Co"), class = "factor"), Quarter = structure(c(2L, 
                                                                                                                                                               3L, 4L, 5L, 6L, 1L, 2L, 3L, 2L, 3L, 4L, 5L, 6L, 1L, 3L, 4L, 5L, 
                                                                                                                                                               6L), .Label = c("1Q 2013", "1Q 2014", "1Q 2015", "2Q 2014", "3Q 2014", 
                                                                                                                                                                               "4Q 2014"), class = "factor"), Revenue = c(5000L, 6000L, 3200L, 
                                                                                                                                                                                                                          1200L, 7200L, 10000L, 2500L, 4100L, 1250L, 8100L, 2300L, 3700L, 
                                                                                                                                                                                                                          1100L, 1600L, 8000L, 9000L, 10000L, 12000L)), .Names = c("Company", 
                                                                                                                                                                                                                                                                                   "Quarter", "Revenue"), class = "data.frame", row.names = c(NA, 
                                                                                                                                                                                                                                                                                                                                              -18L))

第一个是直截了当的，但使用正则表达式来分隔季度和年份

(x <- gsub('(..)(?:\\s+)(\\d{4})', '\\1', dat$Quarter))
# [1] "1Q" "1Q" "2Q" "3Q" "4Q" "1Q" "1Q" "1Q" "1Q" "1Q" "2Q" "3Q" "4Q" "1Q" "1Q"
# [16] "2Q" "3Q" "4Q"

(y <- gsub('(..)(?:\\s+)(\\d{4})', '\\2', dat$Quarter))
# [1] "2014" "2015" "2014" "2014" "2014" "2013" "2014" "2015" "2014" "2015"
# [11] "2014" "2014" "2014" "2013" "2015" "2014" "2014" "2014"

然后按组合排序：

dat[order(y, x), ]
#     Company Quarter Revenue
# 6      ACME 1Q 2013   10000
# 14  ABC Inc 1Q 2013    1600
# 1      ACME 1Q 2014    5000
# 7   ABC Inc 1Q 2014    2500
# 9   ABC Inc 1Q 2014    1250
# 3      ACME 2Q 2014    3200
# 11  ABC Inc 2Q 2014    2300
# 16 Handy Co 2Q 2014    9000
# 4      ACME 3Q 2014    1200
# 12  ABC Inc 3Q 2014    3700
# 17 Handy Co 3Q 2014   10000
# 5      ACME 4Q 2014    7200
# 13  ABC Inc 4Q 2014    1100
# 18 Handy Co 4Q 2014   12000
# 2      ACME 1Q 2015    6000
# 8   ABC Inc 1Q 2015    4100
# 10  ABC Inc 1Q 2015    8100
# 15 Handy Co 1Q 2015    8000

第二个在数据框中再创建两列，而不是像通常那样创建向量和顺序：

(tmp <- data.frame(do.call('rbind', strsplit(as.character(dat$Quarter), ' ')),
                  stringsAsFactors = FALSE))
#    X1   X2
# 1  1Q 2014
# 2  1Q 2015
# 3  2Q 2014
# 4  3Q 2014
# 5  4Q 2014
# 6  1Q 2013
# 7  1Q 2014
# 8  1Q 2015
# 9  1Q 2014
# 10 1Q 2015
# 11 2Q 2014
# 12 3Q 2014
# 13 4Q 2014
# 14 1Q 2013
# 15 1Q 2015
# 16 2Q 2014
# 17 3Q 2014
# 18 4Q 2014

dat[order(tmp[, 2], tmp[, 1]), ]

#     Company Quarter Revenue
# 6      ACME 1Q 2013   10000
# 14  ABC Inc 1Q 2013    1600
# 1      ACME 1Q 2014    5000
# 7   ABC Inc 1Q 2014    2500
# 9   ABC Inc 1Q 2014    1250
# 3      ACME 2Q 2014    3200
# 11  ABC Inc 2Q 2014    2300
# 16 Handy Co 2Q 2014    9000
# 4      ACME 3Q 2014    1200
# 12  ABC Inc 3Q 2014    3700
# 17 Handy Co 3Q 2014   10000
# 5      ACME 4Q 2014    7200
# 13  ABC Inc 4Q 2014    1100
# 18 Handy Co 4Q 2014   12000
# 2      ACME 1Q 2015    6000
# 8   ABC Inc 1Q 2015    4100
# 10  ABC Inc 1Q 2015    8100
# 15 Handy Co 1Q 2015    8000