Working with dates using `lubridate`

Author

Emily Malcolm-White

Date Formats

Think of how many different formats you know of to format a date:

2023 07 06
Wed, Jun 7, 2023
07-06-23
06-07-23 14:55 ET
06/07/2023 2:55pm

Yikes!

Date, Time, and Datetime

Date/time data are data that conveys information about, you guessed it, date and/or time! There are three relevant data types when we talk about date/time data:

Date - only has the date (e.g. 2020-05-15)
Time - only has the time (e.g. 20:45:00)
Datetime - has both the date and time (e.g. 2020-05-15 20:45:00)

`Lubridate`

#LOAD PACKAGES 
library(tidyverse)
library(lubridate)

Standard Date Format

The ymd() function transforms data in all kinds of different formats into a standardized date format displaying year, then month, then day.

ymd("06 02 04")

[1] "2006-02-04"

ymd("06/02/04")

[1] "2006-02-04"

ymd("20060204")  # works as well

[1] "2006-02-04"

ymd("2006 2 4")

[1] "2006-02-04"

ymd(060204)  # works with numbers

[1] "2006-02-04"

mdy() (month day year) and dmy() (day month year) formats also exist.

ymd_hms("2020-04-01 10:30:13")

[1] "2020-04-01 10:30:13 UTC"

ymd_hm("2020/04/01 10.30")

[1] "2020-04-01 10:30:00 UTC"

Solar Data

Shoal Marine Lab (SML) is a remote field station located on Appledore Island, Maine jointly operated by Cornell University and the University of New Hampshire. The island is powered primarily by solar power.

#read in Data
power <- read.csv("data/power.csv")

What format is the date in?

Warning

What happens if we try to make a line plot with the date in this format?

power %>% 
  ggplot(aes(x=Date, y=power_kW)) +
  geom_line()

Yikes!

We need to put it in standardized date format first:

power <- power %>% 
  mutate(Date = mdy_hm(Date))

Standardized Format helps us to create time series plots very easily!

power %>% 
  ggplot(aes(x=Date, y=power_kW)) +
  geom_line()

Picking out information

Sometimes we need to pick out year, month, date so we can filter, sort, etc.

power <- power %>% 
  mutate(Year = year(Date)) %>%   
  mutate(Month = month(Date)) %>% 
  mutate(Day = day(Date))

Suppose we only want the time series plot for June 1:

power %>% 
  filter(Month == "6") %>%
  filter(Day == "1") %>% 
  ggplot(aes(x=Date, y=power_kW)) +
  geom_line()

Portal Data

The Portal Project is a long-term ecological study being conducted near Portal, AZ. Since 1977, the site has been used to study the interactions among rodents, ants and plants and their respective responses to climate.

#LOAD DATA
portal_rodent <- read.csv("https://github.com/weecology/PortalData/raw/main/Rodents/Portal_rodent.csv")

Unfortunately, because the information about datetime is divided up into different columns, R does not recognize it as date/time data. What we need to do is combine and convert all of these columns into datetime. To do this, we can use the function make_datetime()

portal_rodent <- portal_rodent %>% 
  mutate(datetime = make_datetime(year, month, day))

--- title: 'Working with dates using `lubridate`' author: 'Emily Malcolm-White' format: html: toc: TRUE code-overflow: wrap embed-resources: true code-tools: source: true toggle: false caption: none code-annotations: hover pdf: default execute: message: FALSE warning: FALSE --- ![](https://lubridate.tidyverse.org/logo.png){width=30%} # Date Formats Think of how many different formats you know of to format a date: - 2023 07 06 - Wed, Jun 7, 2023 - 07-06-23 - 06-07-23 14:55 ET - 06/07/2023 2:55pm Yikes! # Date, Time, and Datetime Date/time data are data that conveys information about, you guessed it, date and/or time! There are three relevant data types when we talk about date/time data: 1. Date - only has the date (e.g. 2020-05-15) 2. Time - only has the time (e.g. 20:45:00) 3. Datetime - has both the date and time (e.g. 2020-05-15 20:45:00) # `Lubridate` ![Artwork by @allisonhorst](https://cdn.myportfolio.com/45214904-6a61-4e23-98d6-b140f8654a40/79a12c01-0cc1-4643-b1e7-8ce8cafc947b_rw_1920.png?h=a3760f3882e735d7065627fc060714ce){width=50%} ```{r} #| warning: FALSE #| message: FALSE #LOAD PACKAGES library(tidyverse) library(lubridate) ``` # Standard Date Format The `ymd()` function transforms data in all kinds of different formats into a standardized date format displaying year, then month, then day. ```{r} ymd("06 02 04") ymd("06/02/04") ymd("20060204") # works as well ymd("2006 2 4") ymd(060204) # works with numbers ``` `mdy()` (month day year) and `dmy()` (day month year) formats also exist. ```{r} ymd_hms("2020-04-01 10:30:13") ymd_hm("2020/04/01 10.30") ``` # Solar Data Shoal Marine Lab (SML) is a remote field station located on Appledore Island, Maine jointly operated by Cornell University and the University of New Hampshire. The island is powered primarily by solar power. ```{r} #read in Data power <- read.csv("data/power.csv") ``` What format is the date in? :::callout-warning What happens if we try to make a line plot with the date in this format? ```{r} power %>% ggplot(aes(x=Date, y=power_kW)) + geom_line() ``` Yikes! ::: We need to put it in standardized date format first: ```{r} power <- power %>% mutate(Date = mdy_hm(Date)) ``` Standardized Format helps us to create time series plots very easily! ```{r} power %>% ggplot(aes(x=Date, y=power_kW)) + geom_line() ``` # Picking out information ![Artwork by @allisonhorst](https://cdn.myportfolio.com/45214904-6a61-4e23-98d6-b140f8654a40/3017cf57-d7d9-4f73-9b66-f2be0ebf2489_rw_1920.png?h=2cdd179bc087c56b43f8aa6ea849ca1a){width=50%} Sometimes we need to pick out year, month, date so we can filter, sort, etc. ```{r} power <- power %>% mutate(Year = year(Date)) %>% mutate(Month = month(Date)) %>% mutate(Day = day(Date)) ``` Suppose we only want the time series plot for June 1: ```{r} power %>% filter(Month == "6") %>% filter(Day == "1") %>% ggplot(aes(x=Date, y=power_kW)) + geom_line() ``` # Portal Data The Portal Project is a long-term ecological study being conducted near Portal, AZ. Since 1977, the site has been used to study the interactions among rodents, ants and plants and their respective responses to climate. ```{r} #LOAD DATA portal_rodent <- read.csv("https://github.com/weecology/PortalData/raw/main/Rodents/Portal_rodent.csv") ``` Unfortunately, because the information about datetime is divided up into different columns, R does not recognize it as date/time data. What we need to do is combine and convert all of these columns into datetime. To do this, we can use the function `make_datetime()` ```{r} portal_rodent <- portal_rodent %>% mutate(datetime = make_datetime(year, month, day)) ```