inst/doc/trip.R

In tripaccess: American Travel Behavior and Access Datasets

## ----include = FALSE----------------------------------------------------------
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

## ----eval = FALSE-------------------------------------------------------------
# devtools::install()

## ----setup, message = FALSE, warning = FALSE----------------------------------
library(tripaccess)
library(tidyverse)

## ----message = FALSE, warning = FALSE, out.width = "70%"----------------------
#> Filtered to shorter trips for a clearer introductory visualization
short_trips <- trip |>
  filter(trip_miles <= 50,
         trip_duration <= 180)

#> Create readable labels for trip purpose
short_trips <- short_trips |>
  mutate(
    trip_purpose_label = case_when(
      trip_purpose == "shopping_trip" ~ "Shopping",
      trip_purpose == "other_home_based_trip" ~ "Other home-based",
      trip_purpose == "social_recreational_trip" ~ "Social/recreational",
      trip_purpose == "work_trip" ~ "Work",
      trip_purpose == "other_non_home_based_trip" ~ "Other non-home-based"
    )
  )

#> Sort facet_wrap labels
short_trips$trip_purpose_sort_val <- factor(short_trips$trip_purpose_label, levels = c("Shopping", "Other home-based", "Social/recreational", "Work", "Other non-home-based"))

#> Plot Trip Distance by Trip Purpose
ggplot(data = short_trips,
       aes(x = trip_miles)) +
  geom_histogram(bins = 25, color = "white") +
  facet_wrap(~trip_purpose_sort_val) +
  labs(title = "Trip Distance by Trip Purpose",
       x = "Trip Distance in Miles",
       y = "Count of Trips") +
  theme_bw() +
  theme(strip.text = element_text(size = 3),
        axis.text = element_text(size = 5),
        axis.title = element_text(size = 5),
        title = element_text(size = 5))

#> Summary statistics of trip miles and duration by trip purpose
short_trips |>
  group_by(trip_purpose_label) |>
  summarize(
    trips = n(),
    mean_miles = mean(trip_miles),
    median_miles = median(trip_miles),
    mean_duration = mean(trip_duration),
    median_duration = median(trip_duration)
  ) |>
  arrange(desc(mean_miles))

#> Filtered to trips with positive distance and duration
positive_distance_trips <- short_trips |>
  filter(trip_miles > 0,
         trip_duration > 0)

#> Plot trip duration by trip distance
ggplot(data = positive_distance_trips,
       aes(x = trip_miles,
           y = trip_duration)) +
  geom_point(alpha = 0.05) +
  geom_smooth(method = lm, se = FALSE, formula = y ~ x, color = "blue") +
  labs(title = "Trip Duration by Trip Distance",
       x = "Trip Distance in Miles",
       y = "Trip Duration in Minutes") +
  theme_bw()

#> Fit a simple linear regression model
duration_miles_model <- lm(trip_duration ~ trip_miles,
                           data = positive_distance_trips)
summary(duration_miles_model)

#> Correlation between trip distance and trip duration
cor(positive_distance_trips$trip_miles, positive_distance_trips$trip_duration)

#> The slope estimates the average change in trip duration for one additional
#> mile of trip distance.
coef(duration_miles_model)["trip_miles"]

#> Plot Trip Duration by Trip Purpose
ggplot(data = positive_distance_trips,
       aes(x = trip_duration,
           y = trip_purpose_label,
           fill = trip_purpose_label)) +
  geom_boxplot() +
  labs(title = "Trip Duration by Trip Purpose",
       fill = "Trip Purpose",
       x = "Trip Duration in Minutes",
       y = "Trip Purpose") +
  theme_bw() +
  theme(axis.text = element_text(size = 4),
        axis.title = element_text(size = 4),
        title = element_text(size = 4),
        legend.text = element_text(size = 4),
        legend.title = element_text(size = 4))

#> Test whether average trip distance differs by trip purpose
trip_purpose_model <- lm(trip_miles ~ trip_purpose_label,
                         data = positive_distance_trips)
anova(trip_purpose_model)