Introduction to tsibble

The tsibble package extends the tidyverse to temporal data. Built on top of the tibble, a tsibble (or tbl_ts) is a data- and model-oriented object. Compared to the conventional time series objects in R, for example ts, zoo, and xts, the tsibble preserves time indices as the essential data column and makes heterogeneous data structures possible. Beyond the tibble-like representation, key comprised of single or multiple variables is introduced to uniquely identify observational units over time (index). The tsibble package aims at managing temporal data and getting analysis done in a fluent workflow.

Contextual semantics: index and key

tsibble() creates a tsibble object, and as_tsibble() is an S3 method to coerce other objects to a tsibble. An object that a vector/matrix underlies, such as ts and mts, can be automated to a tsibble using as_tsibble() without any specification. If it is a tibble or data frame, as_tsibble() requires a little more setup in order to declare the index and key variables.

library(dplyr)
library(lubridate)
library(tsibble)
weather <- nycflights13::weather %>% 
  select(origin, time_hour, temp, humid, precip)
weather
#> # A tibble: 26,115 × 5
#>   origin time_hour            temp humid precip
#>   <chr>  <dttm>              <dbl> <dbl>  <dbl>
#> 1 EWR    2013-01-01 01:00:00  39.0  59.4      0
#> 2 EWR    2013-01-01 02:00:00  39.0  61.6      0
#> 3 EWR    2013-01-01 03:00:00  39.0  64.4      0
#> 4 EWR    2013-01-01 04:00:00  39.9  62.2      0
#> 5 EWR    2013-01-01 05:00:00  39.0  64.4      0
#> # ℹ 26,110 more rows

The weather data included in the package nycflights13 contains the hourly meteorological records (such as temperature, humid and precipitation) over the year of 2013 at three stations (i.e. JFK, LGA and EWR) in New York City. Since the time_hour is the only column involving the timestamps, as_tsibble() defaults it to the index variable; alternatively, the index can be specified by the argument index = time_hour to disable the verbose message.

Except for index, a tsibble requires “key”, which defines subjects or individuals measured over time. In this example, the origin variable is the identifier, which is passed to the argument key in as_tsibble(). Each observation should be uniquely identified by index and key in a valid tsibble. Others—temp, humid and precip—are referred to as measured variables. When creating a tsibble, the key will be sorted first, followed by arranging time from past to recent.

weather_tsbl <- as_tsibble(weather, key = origin)
#> Using `time_hour` as index variable.
weather_tsbl
#> # A tsibble: 26,115 x 5 [1h] <America/New_York>
#> # Key:       origin [3]
#>   origin time_hour            temp humid precip
#>   <chr>  <dttm>              <dbl> <dbl>  <dbl>
#> 1 EWR    2013-01-01 01:00:00  39.0  59.4      0
#> 2 EWR    2013-01-01 02:00:00  39.0  61.6      0
#> 3 EWR    2013-01-01 03:00:00  39.0  64.4      0
#> 4 EWR    2013-01-01 04:00:00  39.9  62.2      0
#> 5 EWR    2013-01-01 05:00:00  39.0  64.4      0
#> # ℹ 26,110 more rows

An interval is automatically obtained based on the corresponding time representation:

  • integer/numeric/ordered: either “unit” or “year” (Y)
  • yearquarter/yearqtr: “quarter” (Q)
  • yearmonth/yearmon: “month” (M)
  • yearweek: “week” (W)
  • Date: “day” (D)
  • difftime: “week” (W), “day” (D), “hour” (h), “minute” (m), “second” (s)
  • POSIXct/hms: “hour” (h), “minute” (m), “second” (s), “millisecond” (us), “microsecond” (ms)
  • nanotime: “nanosecond” (ns)

That is, a tsibble of monthly intervals expects the yearmonth/yearmon class in the index column. Neither Date nor POSIXct gives a monthly tsibble.

The print display is data-centric and contextually informative, such as data dimension, time interval, and the number of time-based units. Above displays the weather_tsbl its one-hour interval ([1h]) and the origin [3] as the key along with three time series in the table.

Data pipeline

This tidy data representation most naturally supports thinking of operations on the data as building blocks, forming part of a “data pipeline” in time-based context. Users who are familiar with tidyverse would find it easier to perform common temporal analysis tasks. For example, index_by() is the counterpart of group_by() in temporal context, but it only groups the time index. index_by() + summarise() is used to summarise daily highs and lows at each station. As a result, the index is updated to the date with one-day interval from the index time_hour; two new variables are created and computed for daily maximum and minimum temperatures.

weather_tsbl %>%
  group_by_key() %>%
  index_by(date = ~ as_date(.)) %>% 
  summarise(
    temp_high = max(temp, na.rm = TRUE),
    temp_low = min(temp, na.rm = TRUE)
  )
#> # A tsibble: 1,092 x 4 [1D]
#> # Key:       origin [3]
#>   origin date       temp_high temp_low
#>   <chr>  <date>         <dbl>    <dbl>
#> 1 EWR    2013-01-01      41       28.0
#> 2 EWR    2013-01-02      34.0     24.1
#> 3 EWR    2013-01-03      34.0     26.1
#> 4 EWR    2013-01-04      39.9     28.9
#> 5 EWR    2013-01-05      44.1     32  
#> # ℹ 1,087 more rows

Irregular time interval

Note that the tsibble handles regularly-spaced temporal data well, from seconds to years based on its time representation (see ?tsibble). The option regular, by default, is set to TRUE in as_tsibble(). Specify regular to FALSE to create a tsibble for the data collected at irregular time interval. Below shows the scheduled date time of the flights in New York:

flights <- nycflights13::flights %>%
  mutate(sched_dep_datetime = 
    make_datetime(year, month, day, hour, minute, tz = "America/New_York"))

The key contains columns carrier and flight to identify observational units over time, from a passenger’s point of view. With regular = FALSE, it turns to an irregularly-spaced tsibble, where [!] highlights the irregularity.

flights_tsbl <- flights %>%
  as_tsibble(
    key = c(carrier, flight), 
    index = sched_dep_datetime, 
    regular = FALSE
  )
flights_tsbl
#> # A tsibble: 336,776 x 20 [!] <America/New_York>
#> # Key:       carrier, flight [5,725]
#>    year month   day dep_time sched_dep_time dep_delay arr_time sched_arr_time
#>   <int> <int> <int>    <int>          <int>     <dbl>    <int>          <int>
#> 1  2013    11     3     1531           1540        -9     1653           1725
#> 2  2013    11     4     1539           1540        -1     1712           1725
#> 3  2013    11     5     1548           1540         8     1708           1725
#> 4  2013    11     6     1535           1540        -5     1657           1725
#> 5  2013    11     7     1549           1540         9     1733           1725
#> # ℹ 336,771 more rows
#> # ℹ 12 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
#> #   tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
#> #   hour <dbl>, minute <dbl>, time_hour <dttm>, sched_dep_datetime <dttm>

To regularise an irregular tsibble, it can be achieved with index_by() + summarise().