R语言中POSIXct转Date、字符转数值及日期格式转换问题求助

Hey Drew, let's work through these three data wrangling issues step by step—they’re super common in R, so I’ve got you covered!

If your existing column is already a POSIXct/POSIXt type, converting it to a Date is straightforward with as.Date(), and you can specify the timezone directly to avoid any offset issues:

# Assuming your POSIXct column is named 'time_stamp'
df$date_col <- as.Date(df$time_stamp, tz = "Europe/London")

The reason your earlier dmy_hm() attempt didn’t return a Date type is that dmy_hm() (from the lubridate package) creates a POSIXct object by default—it’s designed for datetime values with hours and minutes, not just dates. Using as.Date() directly on the existing POSIXct column will strip the time component and give you a proper Date type.

Even after replacing --- with NA, hidden characters (like extra spaces, commas instead of periods for decimals, or invisible whitespace) can break the as.numeric() conversion. Here are two reliable fixes:

Option 1: Use readr::parse_number() (most robust) #

The parse_number() function from the readr package automatically ignores non-numeric characters and handles different decimal separators:

# Install readr if you haven't already
# install.packages("readr")
library(readr)
df$value <- parse_number(df$value)

Option 2: Clean the string manually first #

If you prefer base R, use gsub() to remove any characters that aren’t digits, periods, or minus signs (for negative numbers):

# Remove all non-numeric/non-decimal characters
df$value_cleaned <- gsub("[^0-9.-]", "", df$value)
# Convert to numeric
df$value <- as.numeric(df$value_cleaned)

To debug why the original conversion failed, run warning(as.numeric(as.character(df$value)))—this will show you exactly which values are causing errors, so you can target those for cleaning.

Remember that a Date type in R stores dates as integers under the hood—its "display format" is separate from its data type. If you want to display dates as dd/mm/yyyy (as a string) while keeping a Date type for calculations, use format():

# First, make sure you have a proper Date column (from step 1)
df$date_col <- as.Date(df$time_stamp, tz = "Europe/London")
# Create a formatted string column in dd/mm/yyyy
df$date_formatted <- format(df$date_col, "%d/%m/%Y")

If you accidentally created a POSIXct column with dmy_hm(), just convert it to Date first, then apply the format:

# If your 'date' column is POSIXct from dmy_hm()
df$date_col <- as.Date(df$date, tz = "Europe/London")
df$date_formatted <- format(df$date_col, "%d/%m/%Y")

Note: date_formatted will be a character column, which is fine for reporting or exporting. If you need to keep it as a Date type but want R to display it in dd/mm/yyyy by default, you can set a global option (though this affects all Date columns in your session):

options(date.format = "%d/%m/%Y")

内容的提问来源于stack exchange，提问作者Drew