R与Lubridate：导入CSV文件时如何设置时区

报错原因 #

col_datetime()函数本身不支持locale或tz参数，这些参数需要传递给read_csv()的全局设置，或者在导入后单独处理datetime列。

正确设置时区的两种方法 #

方法1：在read_csv中全局指定locale #

将时区设置放在read_csv()的locale参数中，所有datetime列都会应用这个时区：

example3 <- read_csv("NAME, TRNSCO, ID, DPMNEM, DATE, PEAK
                 BEC,EGSI,1062,SERP,10/12/2022 18:00,6.464
                 BEC,EGSI,1072,SPTR,10/30/2022 22:00,67.456
                 BEC,EGSI,1152,DYNE,10/24/2022 12:00,4.606
                 CLA,ELI,3023,FRMR,10/12/2022 17:00,6.352
                 CLA,ELI,3043,SHON,10/19/2022 8:00,2.296
                 CLA,ELI,3063,BERN,10/20/2022 7:00,5.232",
            col_types = cols_only(
              NAME = col_character(),
              ID =  col_character(),
              DPMNEM = col_character(),
              DATE = col_datetime(format = "%m/%d/%Y %H:%M"),
              PEAK = col_double()),
            locale = locale(tz = "US/Central"))

方法2：导入后单独设置时区 #

先按默认方式导入数据，再用lubridate包的force_tz()函数为datetime列指定时区：

library(lubridate)

# 先导入数据
example3 <- read_csv("NAME, TRNSCO, ID, DPMNEM, DATE, PEAK
                 BEC,EGSI,1062,SERP,10/12/2022 18:00,6.464
                 BEC,EGSI,1072,SPTR,10/30/2022 22:00,67.456
                 BEC,EGSI,1152,DYNE,10/24/2022 12:00,4.606
                 CLA,ELI,3023,FRMR,10/12/2022 17:00,6.352
                 CLA,ELI,3043,SHON,10/19/2022 8:00,2.296
                 CLA,ELI,3063,BERN,10/20/2022 7:00,5.232",
            col_types = cols_only(
              NAME = col_character(),
              ID =  col_character(),
              DPMNEM = col_character(),
              DATE = col_datetime(format = "%m/%d/%Y %H:%M"),
              PEAK = col_double()))

# 设置时区
example3$DATE <- force_tz(example3$DATE, tzone = "US/Central")

提取datetime列的时区信息 #

可以用lubridate包的tz()函数直接提取：

tz(example3$DATE)

或者通过attr()函数获取属性值：

attr(example3$DATE, "tzone")

内容的提问来源于stack exchange，提问作者DavidH