Weather Data

R interface to access the PWS data through the Wunderground API(http://www.wunderground.com/weather/api/d/documentation.html)

To build the package

  R CMD build weatherdata

To check the package

  R CMD check weatherdata_1.0.tar.gz

To install the package

  R CMD install weatherdata_1.0.tar.gz

To run the examples of wunderGround class and pwsVisualization

  require('weatherdata')
  example(wunderGround)
  example(pwsVisualization)

Documentation

  Class "wunderGround"

  Description

  Objects for accessing pws data, pws historical data using the wunderground API

  Usage

    # returns a new object with pws data queried using state city
    ## S4 method for signature 'wunderGround'
  pwsByStateCity(object)

    # returns a new object with pws data queried using lat, long
    ## S4 method for signature 'wunderGround'
  pwsByLatLong(object)

    # returns a new object with pws data queried using country,
    # city
    ## S4 method for signature 'wunderGround'
  pwsByCountryCity(object)

    #returns a data frame with pws data filtered by state, city and distance
    ## S4 method for signature 'wunderGround'
  subtableByStateCityDistance(object,filterState,filterCity,distance)

    #returns a data frame with pws historical data
    ## S4 method for signature 'wunderGround'
  pwsHistoricalData(object,startTime, endTime)
  Arguments

  object   wunderGround object
  filterState      character string e.g. "CA", "UT"
  filterCity       character string e.g. "Moab", "Tahoe","San_Fransisco"
  distance         a numeric object e.g. 5, 6
  startTime        is an object of class POSIXt or POSIXct
  endTime  is an object of class POSIXt or POSIXct
  creating objects of wunderGroundClass, querying the PWS, historical PWS Data

  objects can be created using the new method

  new("wunderGround", state="CA",city="Fremont", distance=5)
  new("wunderGround", latitude=37.776289,longitude =-122.395234)
  new("wunderGround", country="France", city="Paris")
  Note: latitude, longitude, distance are numeric. Also, the distance is an optional field. If distance is not specified the pws data frame will contain all available pws from wunderground.

  If the city has multiple words in it use _ to separate them. For e.g. San_Francisco, Los_Angeles, Santa_Clara

  Also, you can also specify the developer key for accessing the wunderground data through the API.The default developer key used here can do a 10 queries per minute new developer key can be specified using key slot

  To query the PWS:
  using state, city
  obj <- new("wunderGround", state="CA",city="Santa_Clara")
  obj <- pwsByStateCity(obj)
  using country, city
  obj <- new("wunderGround", country="France",city="Paris",distance=5)
  obj <- pwsByCountryCity(obj)
  using latitude, longitude
  obj <- new("wunderGround", latitude=37.776289,longitude=-122.395234)
  obj <- pwsByLatLong(obj)
  The returned object is of type wunderGround and has the PWS data in the pws slot of the object
  Querying historical data:
  obj <- new("wunderGround", state="CA",city="Santa_Clara")
  obj <- pwsByStateCity(obj)
  pwsHistoricalData(obj, startTime=Sys.time() - (3 * 60 *60),endTime=Sys.time())
  pwsHistoricaldata returns a data frame containing the conditions for all pws in obj@pws

  Examples

  obj <- new("wunderGround",state="CA",city="Santa_Clara",distance=5)
  obj <- pwsByStateCity(obj)
  head(obj@pws)

  obj <- new("wunderGround", latitude=37.776289,longitude=-122.395234)
  obj <- pwsByLatLong(obj)
  head(obj@pws)

  obj <- new("wunderGround", country="France",city="Paris",distance=5)
  obj <- pwsByCountryCity(obj)
  head(obj@pws)

  obj <- new("wunderGround", state="CA",city="Santa_Clara")
  obj <- pwsByStateCity(obj)
  subtab <- subtableByStateCityDistance(obj, distance=5)
  head(subtab)

  subtab <- subtableByStateCityDistance(obj, filterCity="Santa Clara")
  head(subtab)
  tab <- pwsHistoricalData(obj, startTime=Sys.time() - (2 * 60 * 60),endTime=Sys.time())
  head(tab)
  tail(tab)

---

  PWS Visualization

  Description

  Methods to plot the PWS historical data

  Usage

    #returns the field descriptions of the PWS historical data
    fieldDescriptions()

    #plot a condition(field) across all PWS
    # This is useful since some of the PWS contain invalid data, we
    # can easily identify those PWS and clean the data
    plotConditionAcrossAllPWS(historicalData, condition)

    #plot condition vs condition for a pws
    plotConditionsForPWS(historicalData, xcondition, ycondition, pws)
  Arguments

  historicalData   data frame containing the PWS data for a given startTime, endTime. See wunderGround class
  xcondition       character string, column name in the data frame,used in the plot for x-axis
  ycondition       character string, column name in the data frame,used in the plot for y-axis
  pws      character string, name of the personal weather station to be filtered from the historical data
  condition        character string, column name in the data frame to be plotted across all pws
  Examples

  obj <- new("wunderGround", state="CA",city="Santa_Clara")
  obj <- pwsByStateCity(obj)
  tab <- pwsHistoricalData(obj, startTime=Sys.time() - (60 * 60),endTime=Sys.time())
  head(tab)
  fieldDescriptions()
  plotConditionAcrossAllPWS(historicalData=tab, condition="tempi")
  plotConditionAcrossAllPWS(historicalData=tab, condition="dewptm")
  plotConditionsForPWS(historicalData=tab,xcondition="tempm",ycondition="dewptm", pws="KCASANTA87")
  plotConditionsForPWS(historicalData=tab, ycondition="tempi",xcondition="observationTime", pws="KCASANTA87")