getInterpolatedDataByCity: Get k-nearest station data, Filtered, Interpolated

In mpiccirilli/weatheR: Download, Transform, and plot NOAA ISD weather data

Description

This function applies four filters: 1) Remove stations with little to no data 2) Remove stations that exceed a maximum distance from each city's reference point 3) Remove stations that exceed a threshold of missing data, including NA values 4) Select closest station remaining for each city, as all remaining stations are deemed adequate

Usage

getInterpolatedDataByCity(city.list, station.list, k = 5, begin, end,
  distance = 100, hourly_interval = 3, tolerance = 0.05)

Arguments

city.list	City of list of Cities. The format should be as follows: "City, State", or "City, Country"
station.list	Full list of ISD stations included in the package
k	The number of stations to return
begin	Start year (4 digits)
end	End year (4 digits)
distance	Maximum distance allowable from each city's reference point
hourly_interval	Minimum hourly interval allowable (1=hourly; 3 = every 3 hours; 6 = every 6 hours, etc..)
tolerance	This is the percent, in decimals, of missing data you will allow. (.05 = 5% of total data)

Details

It then performs two steps to interpolate missing values: 1) Average over all data points in original dataset to find average hourly observations 2) Linearly interpolate hourly data points for missing observations

Value

Returns a single dataframe with hourly observations (including interpolated) of every city.

Examples

## Not run: 
data(stations)
cities <- c("Nairobi, Kenya", "Tema, Ghana", "Accra, Ghana", "Abidjan, Ivory Coast")
hourly.data <- getInterpolatedDataByCity(cities, station.list, 5, 2010, 2013, 100, 3, .05)
dim(hourly.data)

## End(Not run)

mpiccirilli/weatheR documentation built on May 23, 2019, 6:28 a.m.