# quickerstats

## Introduction

The USDA’s National Agricultural Statistics Service collects a wide variety of data on US agriculture. They maintain a web tool (Quick Stats), and an API for accessing the data. While powerful and flexible, both of the existing tools for pulling data can be cumbersome for many users. The goal of quickerstats is to functionality for pulling data that is easier to use. The package does this in two ways:

• Provide a search tool. An important parameter of the Quick Stats is ‘short_desc’ (called Data Item in the web tool), which is a concatenation six other parameters (commodity_desc, class_desc, prodn_practice_desc, util_practice_desc, statisticcat_desc, and unit_desc). A typical workflow would be to drill down into each of these categories to build the desired short_desc string. The problem is that the available options for each successive parameter depend on the which values have already been chosen. Therefore, it is easiest to simply search the available short_desc strings for what you want. Since there are so many short_desc options, quickerstats provides a search function so you can find what you want.

• Provide easy to use functions for pulling state and county data. Most users will be pulling state or county-level data. We streamline the workflow by providing functionality for only these two options (using FIPS codes). Users can access a particular data item for a single state or all states. Additionally, users can access a data item for all counties in the country, all counties in a state, or a single county.

## Installation and setup

devtools::install_github('anderaa/quickerstats', force=TRUE, build_vignettes=TRUE)
library('quickerstats')

You will need an API key from NASS. Go to https://quickstats.nass.usda.gov/api. I recommend storing your key as an environmental variable. To do so, in your console type:

file.edit("~/.Renviron")

And then add a line to this file:

NASS_KEY='your_nass_key'

Save the file, and restart your R session. You can now access the variable with:

key <- Sys.getenv('NASS_KEY')

## Search for a data item

Search for a data item like this:

items <- search_data_items(key=key, search_terms=c('corn', 'harvested'), exclude=c('sweet'))
print(items[1:5])

This will give you a list of data items. Each will contain all of the search terms and none of the exclude terms. The exclude argument is optional. A typical workflow would be to start broad, view results, then incrementally narrow your search using both arguments. Note that the search terms are not case sensitive - the function handles this automatically.

You can also retrieve options associated with a data item:

df <- get_options(key=key, data_item='CORN, GRAIN - ACRES HARVESTED')
print(head(df, 5))

## Retrieve data

Once you have found your desired data item, you can pull data in one of five ways:

#### 1. Pull state-level data for all states

To pull data for all states, the fips argument must be set to ‘all’:

df <- get_state_data(key=key, year=2017, data_item='CORN, GRAIN - ACRES HARVESTED', fips='all')
print(head(df, 5))

#### 2. Pull state-level data for a single state

To pull data for a single county, the fips argument must be passed a 2-character string that is the state FIPS code:

df <- get_state_data(key=key, year=2017, data_item='CORN, GRAIN - ACRES HARVESTED', fips='08')
print(head(df, 5))

#### 3. Pull county-level data for all counties in the US

As in the state example, set fips to ‘all’:

df <- get_county_data(key=key, year=2017, data_item='CORN, GRAIN - ACRES HARVESTED', fips='all')
print(head(df, 5))

#### 4. Pull county-level data for all counties in a single state

To get data for counties in a state, fips must be the 2-character fips code of the desired state:

df <- get_county_data(key=key, year=2017, data_item='CORN, GRAIN - ACRES HARVESTED', fips='08')
print(head(df, 5))

#### 5. Pull county-level data for a single county

Simple pass fips a 5-character FIPS code for the county:

df <- get_county_data(key=key, year=2017, data_item='CORN, GRAIN - ACRES HARVESTED', fips='08069')
print(head(df, 5))

## Function documentation

?search_data_items
?get_options
?get_county_item_count
?get_county_data
?get_state_item_count
?get_state_data