In byu-transpolab/nhts2017: National Household Travel Survey Data

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

nhtsr

The goal of nhtsr is to make it considerably easier for R users to interact with NHTS 2017 and 2022 datasets. The package contains eight datasets:

• nhts_households and nhts22_households
• nhts_persons and nhts22_persons
• nhts_vehicles and nhts22_vehicles
• nhts_trips and nhts22_trips

Citation:

From ORNL website:

To recognize the valuable role of National Household Travel Survey (NHTS) data in the transportation research process and to facilitate repeatability of the research, users of NHTS data are asked to formally acknowledge the data source. Where possible, this acknowledgment should take place in the form of a formal citation, such as when writing a research report, planning document, on-line article, and other publications. The citation can be formatted as follows:

U.S. Department of Transportation, Federal Highway Administration, 2022
National Household Travel Survey. URL: http://nhts.ornl.gov.

Installation

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("byu-transpolab/nhts2017")

Example

Each of the datasets is a properly data-typed tibble, derived from the SPSS files distributed by Oak Ridge National Laboratory. The variables have attribute labels that appear in RStudio's data set viewer, and factor variables have correct labels appended.

For instance, to count the number of households completing records for each day, we can simply do

library(nhtsr)
library(haven)
library(tidyverse)

nhts_households %>%
  group_by(travday) %>%
  summarise(
    count = n(),
    weighted = sum(wthhfin)
  )

In one departure from the NHTS public data files, the datasets are tidy in that each field appears only once in the dataset. E.g., the msasize variable --- indicating the size of the metropolitan area each household resides in --- is only appended to the nhts_households tibble rather than to all four tibbles. Joining is trivial, however.

nhts22_trips |> 
  left_join(nhts22_households, by = "houseid") |> 
  group_by(msasize) |> 
  summarise(
    mean_trip_length = weighted.mean(trpmiles, wttrdfin)
  )

Additionally, the strttime and endtime fields on the trips data have been converted from four-character strings (e.g. 1310 for 1:10 PM) into R datetime objects. This required setting a date, which was arbitrarily chosen to be an appropriate weekday in October 2017 or October 2022

ggplot(nhts_trips, aes(x = strttime)) + 
  geom_histogram()

byu-transpolab/nhts2017 documentation built on Oct. 10, 2024, 8:48 p.m.