Foundations of
Data Science

Spring 2026 | Data 2 (399)
Jeffrey M. Girard | Lecture 12c

Roadmap

• Parse strings and build dates from individual components

• Extract, round, and modify date-time components

• Understand periods, durations, and intervals for time spans

Creating Dates & Times

What are Dates and Times?

• Question: Why not just use strings for dates?

• Dates and Times are special data types that understand the rules of the calendar and the clock.

• Unlike strings, they know that leap years exist, how many days are in each month, and how to do math with time.

• We use the {lubridate} package (part of the {tidyverse}) to work with them.
• There are three main types to know: a date, a time, and a date-time (which contains both).

Current Date and Time

Question: How can I get the current date or time in R?

# Get the current date
today()

[1] "2026-04-06"

# Get the current date AND time
now()

[1] "2026-04-06 14:13:39 CDT"

Note: R defaults to the YYYY-MM-DD format (the ISO 8601 international standard). This completely avoids the ambiguity between US and European date formats and ensures that dates sort chronologically even when they are treated as plain text!

Parsing Dates from Strings

Question: How do I convert a character string into a proper Date object?

We can build them from strings using the matching {lubridate} function.

# Year, Month, Day
ymd("2026-04-06")

[1] "2026-04-06"

# Month, Day, Year (even handles suffixes and commas)
mdy("April 6th, 2026")

[1] "2026-04-06"

# Day, Month, Year
dmy("06-Apr-2026")

[1] "2026-04-06"

Parsing Date-Times

Question: What if my string also includes the time?

ymd_hm("2026-04-06 16:00") # assumes SS=00 and UTC time zone

[1] "2026-04-06 16:00:00 UTC"

ymd_hms("2026-04-06 16:00:42") # assumes UTC time zone

[1] "2026-04-06 16:00:42 UTC"

ymd_hms("2026-04-06 16:00:42", tz = "US/Central") # fully explicit

[1] "2026-04-06 16:00:42 CDT"

Note: You can view all the time zone names using the OlsonNames() function

Building from Components

Sometimes your data has the date components spread across multiple columns. We will use the flights dataset to see this in action.

data("flights", package = "nycflights13")

flights_dt <- flights |> 
  select(year, month, day, hour, minute, dep_delay) |> 
  print()

# A tibble: 336,776 × 6
   year month   day  hour minute dep_delay
  <int> <int> <int> <dbl>  <dbl>     <dbl>
1  2013     1     1     5     15         2
2  2013     1     1     5     29         4
3  2013     1     1     5     40         2
4  2013     1     1     5     45        -1
5  2013     1     1     6      0        -6
6  2013     1     1     5     58        -4
# ℹ 336,770 more rows

Making Dates and Date-Times

Question: How do I combine separate columns into one date variable?

flights_dt <- flights_dt |> 
  mutate(
    date = make_date(year, month, day),
    datetime = make_datetime(year, month, day, hour, minute)
  ) |> 
  print(n = 5)

# A tibble: 336,776 × 8
   year month   day  hour minute dep_delay date       datetime           
  <int> <int> <int> <dbl>  <dbl>     <dbl> <date>     <dttm>             
1  2013     1     1     5     15         2 2013-01-01 2013-01-01 05:15:00
2  2013     1     1     5     29         4 2013-01-01 2013-01-01 05:29:00
3  2013     1     1     5     40         2 2013-01-01 2013-01-01 05:40:00
4  2013     1     1     5     45        -1 2013-01-01 2013-01-01 05:45:00
5  2013     1     1     6      0        -6 2013-01-01 2013-01-01 06:00:00
# ℹ 336,771 more rows

Working with Components

Pulling out Pieces

Question: How do I get just the month or day out of a date-time?

# Start with a single date-time
my_datetime <- ymd_hms("2016-07-08 12:34:56")

year(my_datetime) # Year

[1] 2016

month(my_datetime) # Month (number)

[1] 7

mday(my_datetime) # Day of the month

[1] 8

Getting Labels

Question: Can I get the actual name of the month or weekday?

Yes, set label = TRUE to return a factor instead of a number.

# Get the month label
month(my_datetime, label = TRUE)

[1] Jul
12 Levels: Jan < Feb < Mar < Apr < May < Jun < Jul < Aug < Sep < ... < Dec

# Get the weekday label
wday(my_datetime, label = TRUE)

[1] Fri
Levels: Sun < Mon < Tue < Wed < Thu < Fri < Sat

Visualizing Weekdays

Extracting components is incredibly useful for summaries and plots.

# Which weekdays had the most flights?
flights_dt |> 
  mutate(
    weekday = wday(date, label = TRUE),
    weekday = fct_infreq(weekday)
  ) |> 
  ggplot(aes(x = weekday)) + 
  geom_bar()

Summarizing by Time

# Calculate average departure delay per weekday
flights_dt |> 
  mutate(weekday = wday(date, label = TRUE)) |> 
  summarize(
    .by = weekday,
    m_delay = mean(dep_delay, na.rm = TRUE)
  ) |> 
  arrange(m_delay) |> 
  print(n = 7)

# A tibble: 7 × 2
  weekday m_delay
  <ord>     <dbl>
1 Sat        7.65
2 Tue       10.6 
3 Sun       11.6 
4 Wed       11.8 
5 Fri       14.7 
6 Mon       14.8 
7 Thu       16.1 

Rounding Dates

Question: How can I group dates together by week or month?

# Round down to the nearest week
floor_date(my_datetime, unit = "week")

[1] "2016-07-03 UTC"

# Round to the nearest month
round_date(my_datetime, unit = "month")

[1] "2016-07-01 UTC"

# Round up to the nearest year
ceiling_date(my_datetime, unit = "year")

[1] "2017-01-01 UTC"

Visualizing Rounded Dates

Rounding dates allows us to aggregate and plot data over time. Notice that the x-axis includes the year, which is redundant since it’s always 2013!

flights_dt |> 
  mutate(
    month = floor_date(
      date, 
      unit = "month"
    )
  ) |> 
  ggplot(aes(x = month)) + 
  geom_bar()

Formatting Date Axes

Question: How can I format the date labels on a plot’s axis?

Use scale_x_date() with the date_labels argument. You can use special % codes to format the dates (e.g., %b for abbreviated month names).

flights_dt |> 
  mutate(
    month = floor_date(
      date, 
      unit = "month"
    )
  ) |> 
  ggplot(aes(x = month)) + 
  geom_bar() +
  scale_x_date(date_labels = "%b") 

Common Date Formats

When formatting dates for plots or parsing tricky strings, you will use special % codes to represent different calendar components.

Code	Meaning	Example
%Y	4-digit year	“2026”
%y	2-digit year	“26”
%m	Month as a number	“04”
%B	Full month name	“April”
%b	Abbreviated month	“Apr”
%d	Day of the month	“06”

Time Spans

Three Types of Time Spans

When doing math with time, R provides three different classes of time spans. Picking the right one depends on what you are trying to calculate.

1. Periods: Represent “human” units of time like weeks and months.
2. Durations: Represent an exact mathematical span of seconds.
3. Intervals: Represent a specific span between a start and end point.

Periods

Periods use human logic, e.g., if you add one month to Jan 15, you get Feb 15th

# Periods use plural functions
days(7)

[1] "7d 0H 0M 0S"

months(2)

[1] "2m 0d 0H 0M 0S"

# We can add periods to a starting date
today() + days(7)

[1] "2026-04-13"

Durations

Durations use strict math. They are always tracked in exact seconds. A duration year is exactly 365.25 days of seconds.

# Durations are prefixed with a 'd'
ddays(7)

[1] "604800s (~1 weeks)"

dyears(1)

[1] "31557600s (~1 years)"

# Because they are exact seconds, you can easily do math with them
2 * dyears(1)

[1] "63115200s (~2 years)"

The DST Trap

Question: Why do we need both Periods and Durations?

Because human clocks are messy! Consider Daylight Saving Time.

# 1 AM on the day we "spring forward" in the US
one_am <- ymd_hms("2026-03-08 01:00:00", tz = "America/New_York")

# Adding a Period respects human clocks (we stay at 1 AM the next day)
one_am + days(1)

[1] "2026-03-09 01:00:00 EDT"

# Adding a Duration adds exactly 86,400 seconds (shifts us to 2 AM)
one_am + ddays(1)

[1] "2026-03-09 02:00:00 EDT"

Intervals

Question: How do I calculate the exact time span between two dates?

An interval represents a span of time between two specific points. R handles all the complex leaps and bounds of the calendar for you.

# Create an interval from a start date to an end day
my_interval <- interval(start = ymd("1987-05-30"), end = today())
my_interval

[1] 1987-05-30 UTC--2026-04-06 UTC

# Divide the interval by a duration to get a number
my_interval / years(1)

[1] 38.85205

Interval Math

You can divide an interval by different units depending on what you want.

# How many months has it been?
my_interval / months(1)

[1] 466.2258

# How many days has it been?
my_interval / days(1)

[1] 14191

# How many hours has it been?
my_interval / hours(1)

[1] 340584

Dealing with Just Times

Question: What if I just have a time, not a date?

The {hms} package is great for parsing and doing math with standalone times.

library(hms)

# Parse time strings
time1 <- parse_hms("12:34:56")
time2 <- parse_hms("12:56:34")

# Subtracting times yields a 'difftime' object
time2 - time1

Time difference of 1298 secs

# You can force the unit to be seconds, mins, or hours
difftime(time2, time1, units = "mins")

Time difference of 21.63333 mins