在R中遍历日期列表时输出数字17641的原因咨询

Great question! Let's break down exactly what's happening here.

First, the core concept: in R, Date objects are stored internally as integers. These integers represent the number of days since a specific reference date (called the "origin"). By default, R uses the Unix epoch—1970-01-01—as this origin point.

Why does R do this? #

Storing dates as integers makes date calculations (like adding days, finding the difference between two dates, or sorting) extremely efficient. Instead of dealing with complex string parsing for every operation, R can rely on simple arithmetic with the underlying integer values.

When do you see the numeric value instead of a formatted date? #

Most of the time, R will display Date objects in a human-readable string format (like 2017-03-08). But you might see the raw numeric value in a few scenarios:

• If you explicitly convert the Date to a numeric type using as.numeric().
• If there’s an accidental type coercion (e.g., combining Date objects with non-Date objects in a vector, though R usually tries to preserve the Date class when possible).
• If there’s a mistake in your date parsing code—like the typo in your snippet—that leads to unintended date interpretation.

Looking at your code:

test_list <- c(as.Date('2017-03-08','2017-03-08' )) 
for(el in test_list){ print(el) }

The issue here is that the second argument to as.Date() is supposed to be a format string (like "%Y-%m-%d"), not another date string. When you pass '2017-03-08' as the format, R tries to parse the first string using that literal format, which leads to an unexpected date result. The numeric value 17641 you see corresponds to 2018-05-20—that’s exactly 17641 days after the 1970-01-01 origin.

Corrected code example #

To create a vector of the 2017-03-08 date correctly, use:

# Correct way to make a vector of Date objects
test_list <- rep(as.Date('2017-03-08'), 2)
for(el in test_list){ print(el) }

Running this will print the formatted date 2017-03-08 twice, as expected.

How to verify the numeric ↔ date relationship #

You can easily convert between the numeric and date formats to confirm:

# Get numeric value of a specific date
as.numeric(as.Date("2018-05-20")) # Returns 17641

# Convert numeric back to a date
as.Date(17641, origin = "1970-01-01") # Returns "2018-05-20"

So to wrap up: R uses integers for dates under the hood to keep operations fast and simple. The numeric output you encountered was likely due to a typo in your date parsing code, which caused R to interpret the date incorrectly and show the underlying integer value.

内容的提问来源于stack exchange，提问作者user1700890