update_weather: Update weather data with new data from wunderground

In CraigMohn/wunderWrapper: Wrapper for accessing/storing historic weather data counts using package rwunderground

Description

update_weatherdata Fetch the necessary data from wunderground, archive it, process it and update the hourly detail and the daily summary data frames.

Usage

update_weather(weatherpair, weatherstation, beginDate = NA, endDate = NA,
  updateDir = "", newfileok = FALSE, nofetchlimit = FALSE, ...)

Arguments

weatherpair	a 2 item list containing the hourly and daily dataframes
weatherstation	string containing the station ID (preface zip codes with ZIPxxxxx)
beginDate	a string "YYYYMMDD" identifying the first date to fetch in the interval. If not specified, it chooses the last date specied for the weatherstation in th weatherpair dataframes. If there is no such date, the beginDate is set to the same value as endDate.
endDate	a string "YYYMMDD" identifying the last date to fetch in the interval. If not specified, the current Sys.Date() is used.
updateDir	string containing the locally-archived wunderground data updates directory, so that data fetched can be re-used.
newfileok	if no existing update file found, create one
nofetchlimit	do not retrieve more than 50 unless true
...	validity limit variables for data cleaning

Details

The daily summaries are slightly dependent on the last observation from the preceding day and the first observation from the following day, as midnight values are interpolated and included in the daily measures. The daily measures for the last day observed are cumulative, and change as more observations are included.

Value

a list containing two dataframes. One contains approximately hourly measures of temperature, precipitation, wind, etc, along with time and weatherstation id. The other contains daily mins, max, means and totals for the things in the hourly reports. Mins and maxs are based on all daily observations plus interpolated values for midnight (local time) at the beginning and end of the day. Means are based on data plus the midnight interpolations, weighted by the elapsed time between them. If the data is for a time and place where Daylight Savings Time is in effect, there will be one 23 hour and one 25 hour day per year. Precip totals are based on the average hourly rainfall rate and a 24 hour day.

See Also

rebuild_weather clean_hourly_data

CraigMohn/wunderWrapper documentation built on May 15, 2019, 5:34 p.m.