r语言 - 保持站点标识符前导为零的数字

Question

我有一个来自美国地质调查局(USGS)网站的数据框架，该网站在其站点标识符(site_no)中使用前导零。为了做进一步的分析，我需要保持site_no的前导零;问题是site_no需要是类数字而不是字符。下面是一个模拟问题的示例数据帧。

df1 <- data.frame(matrix(ncol = 3, nrow = 20))
x <- c("site_no", "Date", "Wtemp")
colnames(df1) <- x
df1$site_no <- c("0101","0101","0101","0101","0101",
"0102","0102","0102","0102","0102",
"0103","0103","0103","0103","0103",
"104","104","104","104","104")
df1$Date <- rep(seq(from = as.Date("2020-01-01"), to = as.Date("2020-01-05"), by = 1),4)  
df1$Wtemp <- c(10,NA,NA,NA,15,
20,NA,NA,10,16,
2,4,6,8,10,
12,14,16,18,20)

我尝试了以下操作，但没有成功

df1$site_no <- as.numeric(df1$site_no) # makes numeric but drops the leading zero which needs to be maintained

library(dplyr)
df1 <- df1 %>%
mutate(site_no = ifelse(row_number()<=15, paste0("0", site_no), site_no)) # returns the leading zero but makes class character

Answer 1

选项为formattable中的comma

library(dplyr)
library(formattable)
out <- df1 %>%
mutate(site_no = comma(as.numeric(site_no), 
width = 4, flag = '0', digits = 0, f = 'd', big.mark = ''))
is.numeric(out$site_no)
#[1] TRUE
out$site_no
#[1] 0101 0101 0101 0101 0101 0102 0102 0102 0102 0102 0103 0103 0103 0103 0103 0104 0104 0104 0104 0104
sum(out$site_no)
#[1] 2050

与产出

out
#   site_no       Date Wtemp
#1     0101 2020-01-01    10
#2     0101 2020-01-02    NA
#3     0101 2020-01-03    NA
#4     0101 2020-01-04    NA
#5     0101 2020-01-05    15
#6     0102 2020-01-01    20
#7     0102 2020-01-02    NA
#8     0102 2020-01-03    NA
#9     0102 2020-01-04    10
#10    0102 2020-01-05    16
#11    0103 2020-01-01     2
#12    0103 2020-01-02     4
#13    0103 2020-01-03     6
#14    0103 2020-01-04     8
#15    0103 2020-01-05    10
#16    0104 2020-01-01    12
#17    0104 2020-01-02    14
#18    0104 2020-01-03    16
#19    0104 2020-01-04    18
#20    0104 2020-01-05    20

---

或者选择将列保持为numeric将创建label属性

library(labelled)
out <- df1 %>%
mutate(site_no = as.numeric(site_no), 
site_no = labelled(site_no, setNames(unique(site_no), 
str_pad(unique(site_no), width = 4, pad = '0')))) 
str(out)
#'data.frame':  20 obs. of  3 variables:
# $ site_no: dbl+lbl [1:20] 101, 101, 101, 101, 101, 102, 102, 102, 102, 102, 103, 103, #103, 103, 103, 104, 104, 104, 104, 104
#   ..@ labels: Named num  101 102 103 104
#   .. ..- attr(*, "names")= chr [1:4] "0101" "0102" "0103" "0104"
# $ Date   : Date, format: "2020-01-01" "2020-01-02" "2020-01-03" "2020-01-04" ...
# $ Wtemp  : num  10 NA NA NA 15 20 NA NA 10 16 ...

---

或者我们可以通过使用str_pad来保持character

library(stringr)    
df1 %>%
mutate(site_no = str_pad(site_no, width = 4, pad = '0'))

与产出

#   site_no       Date Wtemp
#1     0101 2020-01-01    10
#2     0101 2020-01-02    NA
#3     0101 2020-01-03    NA
#4     0101 2020-01-04    NA
#5     0101 2020-01-05    15
#6     0102 2020-01-01    20
#7     0102 2020-01-02    NA
#8     0102 2020-01-03    NA
#9     0102 2020-01-04    10
#10    0102 2020-01-05    16
#11    0103 2020-01-01     2
#12    0103 2020-01-02     4
#13    0103 2020-01-03     6
#14    0103 2020-01-04     8
#15    0103 2020-01-05    10
#16    0104 2020-01-01    12
#17    0104 2020-01-02    14
#18    0104 2020-01-03    16
#19    0104 2020-01-04    18
#20    0104 2020-01-05    20

---

或者从base R转换为数字后的sprintf

df1$site_no <- with(df1, sprintf('%04d', as.numeric(site_no)))