The USgas package provides an overview of demand for
natural gas in the US in a time-series format. That includes the
following dataset: * usgas
- The monthly consumption of
natural gas in the US/state level by end-use since 1973 for US level and
1989 for state level. It includes the following end-use categories: -
Commercial Consumption - Delivered to Consumers - Electric Power
Consumption - Industrial Consumption - Lease and Plant Fuel Consumption
- Pipeline Fuel Consumption - Residential Consumption - Vehicle Fuel
Consumption
The package also includes the following datasets, from previous release:
us_total
- The US annual natural gas consumption by
state-level between 1997 and 2019, and aggregate level between 1949 and
2019us_monthly
- The monthly demand for natural gas in the
US between 2001 and 2020us_residential
- The US monthly natural gas residential
consumption by state and aggregate level between 1989 and 2020The us_total
, us_monthly
, and
us_residential
can be derived out of the usgas
dataset. Therefore, those datasets in the process of deprication and
will be removed in the next release to CRAN.
Data source: The US Energy Information Administration API
The usgas
dataset provides a 313 time series focusing on
the consumption of natural gas by end use in the US (aggregated and
state level). It includes the following fields:
date
- A Date, the month and year of the observation
(the day set by default to 1st of the month)}process
- The process type descriptionstate
- The US state namestate_abb
- The US state abbreviationy
- A numeric, the monthly natural gas residential
consumption in a million cubic feetlibrary(USgas)
data("usgas")
str(usgas)
#> 'data.frame': 92783 obs. of 5 variables:
#> $ date : Date, format: "1973-01-01" "1973-01-01" ...
#> $ process : chr "Commercial Consumption" "Residential Consumption" "Commercial Consumption" "Residential Consumption" ...
#> $ state : chr "U.S." "U.S." "U.S." "U.S." ...
#> $ state_abb: chr "U.S." "U.S." "U.S." "U.S." ...
#> $ y : int 392315 843900 394281 747331 310799 648504 231943 465867 174258 326313 ...
#> - attr(*, "units")= chr "MMCF"
#> - attr(*, "product_name")= chr "Natural Gas"
#> - attr(*, "source")= chr "EIA API: https://www.eia.gov/opendata/browser/natural-gas"
head(usgas)
#> date process state state_abb y
#> 1 1973-01-01 Commercial Consumption U.S. U.S. 392315
#> 2 1973-01-01 Residential Consumption U.S. U.S. 843900
#> 3 1973-02-01 Commercial Consumption U.S. U.S. 394281
#> 4 1973-02-01 Residential Consumption U.S. U.S. 747331
#> 5 1973-03-01 Commercial Consumption U.S. U.S. 310799
#> 6 1973-03-01 Residential Consumption U.S. U.S. 648504
The dataset includes the state level and the US aggregate level
(labeled under U.S.
):
unique(usgas$state)
#> [1] "U.S." "Oregon" "Virginia"
#> [4] "Rhode Island" "Arizona" "Washington"
#> [7] "South Dakota" "New Jersey" "Florida"
#> [10] "Alabama" "Louisiana" "Illinois"
#> [13] "Colorado" "New Hampshire" "Maine"
#> [16] "Iowa" "Alaska" "California"
#> [19] "Michigan" "West Virginia" "North Dakota"
#> [22] "Utah" "Pennsylvania" "Missouri"
#> [25] "Montana" "Texas" "Idaho"
#> [28] "Delaware" "South Carolina" "New Mexico"
#> [31] "Massachusetts" "Georgia" "Arkansas"
#> [34] "New York" "Nebraska" "Tennessee"
#> [37] "Indiana" "District of Columbia" "Minnesota"
#> [40] "Wisconsin" "Vermont" "Hawaii"
#> [43] "Wyoming" "Maryland" "Kansas"
#> [46] "Ohio" "Mississippi" "Nevada"
#> [49] "North Carolina" "Oklahoma" "Kentucky"
#> [52] "Connecticut"
In the example below, we will subset and plot the consumption by end-use in the US:
us_agg <- usgas[which(usgas$state == "U.S."),]
head(us_agg)
#> date process state state_abb y
#> 1 1973-01-01 Commercial Consumption U.S. U.S. 392315
#> 2 1973-01-01 Residential Consumption U.S. U.S. 843900
#> 3 1973-02-01 Commercial Consumption U.S. U.S. 394281
#> 4 1973-02-01 Residential Consumption U.S. U.S. 747331
#> 5 1973-03-01 Commercial Consumption U.S. U.S. 310799
#> 6 1973-03-01 Residential Consumption U.S. U.S. 648504
Let’s now use plotly to plot those series:
library(plotly)
plot_ly(data = us_agg,
x = ~ date,
y = ~ y,
color = ~ process,
type = "scatter",
mode = "line") |>
layout(title = "US Monthly Consumption of Natural Gas by End Use",
yaxis = list(title = "MMCF"),
xaxis = list(title = "Source: EIA Website"),
legend = list(x = 0,
y = 0.95))
Similarly, we can subset a couple of states and visualize them. For example, let’s visualize the residential consumption in the New England states. We will start by subsetting the corresponding states in New England, transform the data.frame to wide format, and reorder by date:
ne <- c("Connecticut", "Maine", "Massachusetts",
"New Hampshire", "Rhode Island", "Vermont")
ne_gas <- usgas[which(usgas$state %in% ne),]
head(ne_gas)
#> date process state state_abb y
#> 495 1989-01-01 Commercial Consumption Rhode Island RI 1032
#> 505 1989-01-01 Commercial Consumption New Hampshire NH 842
#> 506 1989-01-01 Commercial Consumption Maine ME 229
#> 523 1989-01-01 Commercial Consumption Massachusetts MA 7394
#> 533 1989-01-01 Commercial Consumption Vermont VT 315
#> 544 1989-01-01 Commercial Consumption Connecticut CT 3909
Next, let’s use the process
column to extract the
residential consumption and plot it:
ne_gas[which(ne_gas$process == "Residential Consumption"),] |>
plot_ly(x = ~ date,
y = ~ y,
color = ~ state,
type = "scatter",
mode = "line") |>
layout(title = "Monthly Residential Consumption of Natural Gas in New England",
yaxis = list(title = "MMCF"),
xaxis = list(title = "Source: EIA Website"),
legend = list(x = 0,
y = 1))