R语言：时间变量与日期的条件操作技术问询

Hey there! Let's work through this problem together. The core challenge here is converting that character-type HMS variable into a format we can actually compare against a cutoff time, then applying your date-shift logic. I'll show you two reliable approaches—one using the popular lubridate package (super user-friendly for time tasks) and another with chron since you mentioned you've looked into that.

Prerequisites #

First, make sure your TimeStamp column is a proper Date type (not character). If it's not, convert it first with:

df$TimeStamp <- as.Date(df$TimeStamp)

Approach 1: Using lubridate (My Go-To) #

lubridate simplifies parsing time strings and doing comparisons. Here's how to implement your logic:

# Load the package
library(lubridate)

# Step 1: Convert character HMS to seconds since midnight (easy to compare numerically)
df$HMS_seconds <- hms(df$HMS) %>% period_to_seconds()

# Step 2: Define our cutoff time (17:30:00) in total seconds
cutoff_1730 <- 17 * 3600 + 30 * 60  # Equals 63000 seconds

# Step 3: Create TimeStampNew with clear conditional logic
df$TimeStampNew <- case_when(
  is.na(df$HMS) ~ df$TimeStamp,  # Keep original date if HMS is missing
  df$HMS_seconds > cutoff_1730 ~ df$TimeStamp + days(1),  # Shift to next day if after 17:30
  TRUE ~ df$TimeStamp  # Keep original date for all other cases
)

Approach 2: Using chron (As You Explored) #

If you prefer sticking with chron, we can convert the character HMS to a times object, which supports direct time comparisons:

# Load the package
library(chron)

# Step 1: Convert character HMS to chron's times object
df$HMS_chron <- times(df$HMS)

# Step 2: Define our cutoff time as a chron times object
cutoff_time <- times("17:30:00")

# Step 3: Create TimeStampNew with nested ifelse()
df$TimeStampNew <- ifelse(
  is.na(df$HMS),
  df$TimeStamp,
  ifelse(df$HMS_chron > cutoff_time, df$TimeStamp + 1, df$TimeStamp)
)

Quick Notes #

• If your HMS strings have inconsistent formatting (e.g., missing leading zeros like "7:30:0" instead of "07:30:00"), lubridate::hms() usually handles this automatically. For chron, you might need to clean the strings first (use stringr::str_pad() to add leading zeros where needed).
• Double-check that TimeStamp is a Date type—if it's a POSIXct datetime, adding days(1) (lubridate) or +1 (base R) will still work, and the core logic stays identical.

内容的提问来源于stack exchange，提问作者Chellarioo