R/weather-norms.R
In aWhereAPI/aWhere-R-Library: API calls for aWhere API
Defines functions
weather_norms_area
weather_norms_latlng
weather_norms_fields
Documented in
weather_norms_area
weather_norms_fields
weather_norms_latlng
#' @title weather_norms_fields
#'
#' @description
#' \code{weather_norms_fields} pulls long term norm weather data from aWhere's API based on field id
#'
#' @details
#' This function allows you to calculate the averages for weather attributes
#' across any range of years for which data are available. The data pulled includes
#' meanTemp, maxTemp, minTemp, precipitation average, solar radiation average,
#' minHumidity, maxHumidity, maxWind and averageWind, along with the standard deviations
#' for these variables. The data pulled is for the field id identified.
#'
#' The data returned in this function
#' allow you to compare this year or previous years to the long-term normals, calculated as
#' the average of those weather conditions on that day in that location over the years specified.
#'
#' Note about dates: The system does not adjust for any difference in dates between the location of the user
#' and where data is being requested from. It is the responsibility of the user to ensure a valid
#' date range is specified given any differences in timezone. These differences can have implications
#' for whether a given date should be requested from the daily_observed functions or the forecast functions
#'
#' @references https://docs.awhere.com/knowledge-base-docs/historical-weather-norms/
#'
#' @param field_id the field_id associated with the location for which you want to pull data.
#' Field IDs are created using the create_field function. (string)
#' @param monthday_start character string of the first month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the start of your date range. e.g. '07-01' (July 1) (required)
#' @param monthday_end character string of the last month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the end of your date range. e.g. '07-01' (July 1) (required)
#' @param year_start character string of the starting year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2008 (required)
#' @param year_end character string of the last year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2015 (required)
#' @param propertiesToInclude character vector of properties to retrieve from API.
#' Valid values are meanTemp, maxTemp, minTemp, precipitation,
#' solar, maxHumidity, minHumidity, dailyMaxWind (optional)
#' @param exclude_year Year or years which you'd like to exclude from
#' your range of years on which to calculate norms. To exclude
#' multiple years, provide a vector of years. You must include
#' at least three years of data with which to calculate the norms. (numeric, optional)
#' @param includeFeb29thData Whether to keep data from Feb 29th on leap years. Because weather/agronomics
#' summary statistics are calculated via the calendar date and 3 years are required
#' to generate a value, data from this date is more likely to be NA. ALlows user
#' to drop this data to avoid later problems (defaults to TRUE)
#' @param keyToUse aWhere API key to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param secretToUse aWhere API secret to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param tokenToUse aWhere API token to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param apiAddressToUse Address of aWhere API to use. For advanced use only. Most users will not need to use this parameter (optional)
#'
#' @import httr
#' @import data.table
#' @import lubridate
#' @import jsonlite
#'
#' @return dataframe of requested data for dates requested
#'
#' @examples
#' \dontrun{weather_norms_fields(field_id = "field_test"
#' ,monthday_start = "06-01"
#' ,monthday_end = "09-01"
#' ,year_start = 2006
#' ,year_end = 2015)}
#' @export
#'
weather_norms_fields <- function(field_id
,monthday_start
,monthday_end
,year_start
,year_end
,propertiesToInclude = ''
,exclude_years = NULL
,includeFeb29thData = TRUE
,keyToUse = awhereEnv75247$uid
,secretToUse = awhereEnv75247$secret
,tokenToUse = awhereEnv75247$token
,apiAddressToUse = awhereEnv75247$apiAddress) {
#Checking Input Parameters
checkCredentials(keyToUse,secretToUse,tokenToUse)
checkValidField(field_id,keyToUse,secretToUse,tokenToUse)
checkNormsStartEndDates(monthday_start,monthday_end)
checkNormsYearsToRequest(year_start,year_end,monthday_start,monthday_end,exclude_years)
checkPropertiesEndpoint('weather_norms',propertiesToInclude)
# Create Logic of API Request
numObsReturned <- 120
calculateAPIRequests <- TRUE
continueRequestingData <- TRUE
yearsToInclude <- setdiff(seq(year_start,year_end,1),exclude_years)
dataList <- list()
# loop through, making requests in chunks of size numObsReturned
while (continueRequestingData == TRUE | calculateAPIRequests == TRUE) {
#If this clause is triggered the progression of API calls will be
#calculated. After each API call the return will be checked for an error
#indicating that the request was too large. If that occurs this loop will
#be reenentered to calculate using the smaller return size
############################################################################
if (calculateAPIRequests == TRUE) {
calculateAPIRequests <- FALSE
#We need to consider whether a year year is present to determine if Fe
#29th data will be returned
includesLeapYear <- any(is.leapyear(yearsToInclude))
if (includesLeapYear == TRUE) {
yearPrefix <- paste0(yearsToInclude[is.leapyear(yearsToInclude)][1],'-')
} else {
yearPrefix <- paste0(yearsToInclude[1],'-')
}
day_start <- ymd(paste0(yearPrefix
,monthday_start))
day_end <- ymd(paste0(yearPrefix
,monthday_end))
temp <- plan_APICalls(day_start
,day_end
,numObsReturned
,includesLeapYear)
allDates <- temp[[1]]
loops <- temp[[2]]
#remove the years
allDates <- gsub(pattern = '20\\d\\d-',replacement = '',x = allDates)
}
#This for loop will make the API requests as calculated from above
############################################################################
for (i in 1:loops) {
starting = numObsReturned*(i-1)+1
ending = numObsReturned*i
monthday_start_toUse <- allDates[starting]
monthday_end_toUse <- allDates[ending]
if(is.na(monthday_end_toUse)) {
tempDates <- allDates[c(starting:length(allDates))]
monthday_start_toUse <- tempDates[1]
monthday_end_toUse <- tempDates[length(tempDates)]
}
##############################################################################
# Create query
urlAddress <- paste0(apiAddressToUse, "/weather")
strBeg <- paste0('/fields')
strCoord <- paste0('/',field_id)
strType <- paste0('/norms')
strMonthsDays <- paste0('/',monthday_start_toUse,',',monthday_end_toUse)
limitString <- paste0('?limit=',numObsReturned)
if (length(exclude_years) != 0) {
strexclude_years <- paste0('&excludeYears=',toString(exclude_years))
} else {
strexclude_years <- ''
}
if (propertiesToInclude[1] != '') {
propertiesString <- paste0('&properties=',paste0(propertiesToInclude,collapse = ','))
} else {
propertiesString <- ''
}
strYearsType <- paste0('/years')
strYears <- paste0('/',year_start,',',year_end)
url <- URLencode(paste0(urlAddress
,strBeg
,strCoord
,strType
,strMonthsDays
,strYearsType
,strYears
,limitString
,strexclude_years
,propertiesString))
doWeatherGet <- TRUE
tryCount <- 0
while (doWeatherGet == TRUE) {
tryCount <- tryCount + 1
postbody = ''
request <- httr::GET(url, body = postbody, httr::content_type('application/json'),
httr::add_headers(Authorization =paste0("Bearer ", tokenToUse)))
a <- suppressMessages(httr::content(request, as = "text"))
temp <- check_JSON(a
,request
,tryCount
,keyToUse
,secretToUse
,tokenToUse)
doWeatherGet <- temp[[1]]
#if the token was updated, this will cause it to be used through function
tokenToUse <- temp[[3]]
#The temp[[2]] will only not be NA when the limit param is too large.
if(!is.na(temp[[2]] == TRUE)) {
numObsReturned <- temp[[2]]
goodReturn <- FALSE
break
} else {
goodReturn <- TRUE
}
rm(temp)
}
if (goodReturn == TRUE) {
#The JSONLITE Serializer properly handles the JSON conversion
x <- jsonlite::fromJSON(a,flatten = TRUE)
data <- data.table::as.data.table(x[[1]])
data <- removeUnnecessaryColumns(data)
dataList[[length(dataList) + 1]] <- data
} else {
#This will break out of the current loop of making API requests so that
#the logic of the API requests can be recalculated
calculateAPIRequests <- TRUE
break
}
}
continueRequestingData <- FALSE
}
data <- unique(rbindlist(dataList))
#Get rid of leap yearData
if (includeFeb29thData == FALSE) {
data <- data[day != '02-29',]
}
currentNames <- data.table::copy(colnames(data))
data[,field_id := field_id]
data.table::setcolorder(data,c('field_id',currentNames))
checkDataReturn_norms(data,monthday_start,monthday_end,year_start,year_end,exclude_years,includeFeb29thData)
return(as.data.frame(data))
}
#' @title weather_norms_latlng
#'
#' @description
#' \code{weather_norms_latlng} pulls long term norm weather data from aWhere's API based on latitude & longitude
#'
#' @details
#' This function allows you to calculate the averages for weather attributes
#' across any range of years for which data are available. The data pulled includes
#' meanTemp, maxTemp, minTemp, precipitation average, solar radiation average,
#' minHumidity, maxHumidity, maxWind and averageWind, along with the standard deviations
#' for these variables. The data pulled is for the latitude & longitude identified.
#'
#' The data returned in this function
#' allow you to compare this year or previous years to the long-term normals, calculated as
#' the average of those weather conditions on that day in that location over the years specified.
#'
#' Note about dates: The system does not adjust for any difference in dates between the location of the user
#' and where data is being requested from. It is the responsibility of the user to ensure a valid
#' date range is specified given any differences in timezone. These differences can have implications
#' for whether a given date should be requested from the daily_observed functions or the forecast functions
#'
#' @references https://docs.awhere.com/knowledge-base-docs/historical-weather-norms-by-geolocation/
#'
#' @param latitude the latitude of the requested location (double, required)
#' @param longitude the longitude of the requested locations (double, required)
#' @param monthday_start character string of the first month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the start of your date range. e.g. '07-01' (July 1) (required)
#' @param monthday_end character string of the last month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the end of your date range. e.g. '07-01' (July 1) (required)
#' @param year_start character string of the starting year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2008 (required)
#' @param year_end character string of the last year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2015 (required)
#' @param propertiesToInclude character vector of properties to retrieve from API.
#' Valid values are meanTemp, maxTemp, minTemp, precipitation,
#' solar, maxHumidity, minHumidity, dailyMaxWind (optional)
#' @param exclude_year Year or years which you'd like to exclude from
#' your range of years on which to calculate norms. To exclude
#' multiple years, provide a vector of years. You must include
#' at least three years of data with which to calculate the norms. (numeric, optional)
#' @param includeFeb29thData Whether to keep data from Feb 29th on leap years. Because weather/agronomics
#' summary statistics are calculated via the calendar date and 3 years are required
#' to generate a value, data from this date is more likely to be NA. ALlows user
#' to drop this data to avoid later problems (defaults to TRUE)
#' @param keyToUse aWhere API key to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param secretToUse aWhere API secret to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param tokenToUse aWhere API token to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param apiAddressToUse: Address of aWhere API to use. For advanced use only. Most users will not need to use this parameter (optional)
#'
#' @import httr
#' @import data.table
#' @import lubridate
#' @import jsonlite
#'
#' @return data.frame of requested data for dates requested
#'
#' @examples
#' \dontrun{weather_norms_latlng(latitude = 39.8282
#' ,longitude = -98.5795
#' ,monthday_start = '02-01'
#' ,monthday_end = '03-10'
#' ,year_start = 2008
#' ,year_end = 2015
#' ,exclude_years = c(2010,2011))}
#' @export
weather_norms_latlng <- function(latitude
,longitude
,monthday_start
,monthday_end
,year_start
,year_end
,propertiesToInclude = ''
,exclude_years = NULL
,includeFeb29thData = TRUE
,keyToUse = awhereEnv75247$uid
,secretToUse = awhereEnv75247$secret
,tokenToUse = awhereEnv75247$token
,apiAddressToUse = awhereEnv75247$apiAddress) {
#Checking Input Parameters
checkCredentials(keyToUse,secretToUse,tokenToUse)
checkValidLatLong(latitude,longitude)
checkNormsStartEndDates(monthday_start,monthday_end)
checkNormsYearsToRequest(year_start,year_end,monthday_start,monthday_end,exclude_years)
checkPropertiesEndpoint('weather_norms',propertiesToInclude)
# Create Logic of API Request
numObsReturned <- 120
calculateAPIRequests <- TRUE
continueRequestingData <- TRUE
yearsToInclude <- setdiff(seq(year_start,year_end,1),exclude_years)
dataList <- list()
# loop through, making requests in chunks of size numObsReturned
while (continueRequestingData == TRUE | calculateAPIRequests == TRUE) {
#If this clause is triggered the progression of API calls will be
#calculated. After each API call the return will be checked for an error
#indicating that the request was too large. If that occurs this loop will
#be reenentered to calculate using the smaller return size
############################################################################
if (calculateAPIRequests == TRUE) {
calculateAPIRequests <- FALSE
#We need to consider whether a year year is present to determine if Fe
#29th data will be returned
includesLeapYear <- any(is.leapyear(yearsToInclude))
if (includesLeapYear == TRUE) {
yearPrefix <- paste0(yearsToInclude[is.leapyear(yearsToInclude)][1],'-')
} else {
yearPrefix <- paste0(yearsToInclude[1],'-')
}
day_start <- ymd(paste0(yearPrefix
,monthday_start))
day_end <- ymd(paste0(yearPrefix
,monthday_end))
temp <- plan_APICalls(day_start
,day_end
,numObsReturned
,includesLeapYear)
allDates <- temp[[1]]
loops <- temp[[2]]
#remove the years
allDates <- gsub(pattern = '20\\d\\d-',replacement = '',x = allDates)
}
#This for loop will make the API requests as calculated from above
############################################################################
for (i in 1:loops) {
starting = numObsReturned*(i-1)+1
ending = numObsReturned*i
monthday_start_toUse <- allDates[starting]
monthday_end_toUse <- allDates[ending]
if(is.na(monthday_end_toUse)) {
tempDates <- allDates[c(starting:length(allDates))]
monthday_start_toUse <- tempDates[1]
monthday_end_toUse <- tempDates[length(tempDates)]
}
##############################################################################
# Create query
urlAddress <- paste0(apiAddressToUse, "/weather")
strBeg <- paste0('/locations')
strCoord <- paste0('/',latitude,',',longitude)
strType <- paste0('/norms')
strMonthsDays <- paste0('/',monthday_start_toUse,',',monthday_end_toUse)
limitString <- paste0('?limit=',numObsReturned)
if (length(exclude_years) != 0) {
strexclude_years <- paste0('&excludeYears=',toString(exclude_years))
} else {
strexclude_years <- ''
}
if (propertiesToInclude[1] != '') {
propertiesString <- paste0('&properties=',paste0(propertiesToInclude,collapse = ','))
} else {
propertiesString <- ''
}
strYearsType <- paste0('/years')
strYears <- paste0('/',year_start,',',year_end)
url <- URLencode(paste0(urlAddress
,strBeg
,strCoord
,strType
,strMonthsDays
,strYearsType
,strYears
,limitString
,strexclude_years
,propertiesString))
doWeatherGet <- TRUE
tryCount <- 0
while (doWeatherGet == TRUE) {
tryCount <- tryCount + 1
postbody = ''
request <- httr::GET(url, body = postbody, httr::content_type('application/json'),
httr::add_headers(Authorization =paste0("Bearer ", tokenToUse)))
a <- suppressMessages(httr::content(request, as = "text"))
temp <- check_JSON(a
,request
,tryCount
,keyToUse
,secretToUse
,tokenToUse)
doWeatherGet <- temp[[1]]
#if the token was updated, this will cause it to be used through function
tokenToUse <- temp[[3]]
#The temp[[2]] will only not be NA when the limit param is too large.
if(!is.na(temp[[2]] == TRUE)) {
numObsReturned <- temp[[2]]
goodReturn <- FALSE
break
} else {
goodReturn <- TRUE
}
rm(temp)
}
if (goodReturn == TRUE) {
#The JSONLITE Serializer properly handles the JSON conversion
x <- jsonlite::fromJSON(a,flatten = TRUE)
data <- data.table::as.data.table(x[[1]])
data <- removeUnnecessaryColumns(data)
dataList[[length(dataList) + 1]] <- data
} else {
#This will break out of the current loop of making API requests so that
#the logic of the API requests can be recalculated
calculateAPIRequests <- TRUE
break
}
}
continueRequestingData <- FALSE
}
data <- unique(rbindlist(dataList))
#Get rid of leap yearData
if (includeFeb29thData == FALSE) {
data <- data[day != '02-29',]
}
currentNames <- data.table::copy(colnames(data))
data[,latitude := latitude]
data[,longitude := longitude]
data.table::setcolorder(data,c('latitude','longitude',currentNames))
checkDataReturn_norms(data,monthday_start,monthday_end,year_start,year_end,exclude_years,includeFeb29thData)
return(as.data.frame(data))
}
#' @title weather_norms_area
#'
#' @description
#' \code{weather_norms_area} pulls long term norm weather data from aWhere's API based on a data.frame of lat/lon, polygon or extentt
#'
#' @details
#' This function allows you to calculate the averages for weather attributes
#' across any range of years for which data are available. The data pulled includes
#' meanTemp, maxTemp, minTemp, precipitation average, solar radiation average,
#' minHumidity, maxHumidity, maxWind and averageWind, along with the standard deviations
#' for these variables. The polygon should be either a SpatialPolygons object or
#' a well-known text character string or an extent.
#'
#' The data returned in this function
#' allow you to compare this year or previous years to the long-term normals, calculated as
#' the average of those weather conditions on that day in that location over the years specified.
#'
#' Note about dates: The system does not adjust for any difference in dates between the location of the user
#' and where data is being requested from. It is the responsibility of the user to ensure a valid
#' date range is specified given any differences in timezone. These differences can have implications
#' for whether a given date should be requested from the daily_observed functions or the forecast functions.
#' Furthermore, because this function can take as input locations that may be in different timezones, it is
#' the responsibility of the user to either ensure that the date range specified is valid for all relevant
#' locations or to break the query into pieces.
#'
#' @references https://docs.awhere.com/knowledge-base-docs/historical-weather-norms-by-geolocation/
#'
#' @param polygon either a data.frame with column names lat/lon, SpatialPolygons object,
#' well-known text string, or extent from raster package. If the object contains
#' multiple polygons, the union of them is used. Information from each individal
#' polygon can be retrieved by returning spatial data and using
#' the over function from the sp package
#' @param monthday_start character string of the first month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the start of your date range. e.g. '07-01' (July 1) (required)
#' @param monthday_end character string of the last month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the end of your date range. e.g. '07-01' (July 1) (required)
#' @param year_start character string of the starting year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2008 (required)
#' @param year_end character string of the last year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2015 (required)
#' @param propertiesToInclude character vector of properties to retrieve from API. Valid values are meanTemp, maxTemp, minTemp, precipitation, solar, maxHumidity, minHumidity, dailyMaxWind (optional)
#' @param exclude_year Year or years which you'd like to exclude from
#' your range of years on which to calculate norms. To exclude
#' multiple years, provide a vector of years. You must include
#' at least three years of data with which to calculate the norms. (numeric, optional)
#' @param includeFeb29thData Whether to keep data from Feb 29th on leap years. Because weather/agronomics
#' summary statistics are calculated via the calendar date and 3 years are required
#' to generate a value, data from this date is more likely to be NA. ALlows user
#' to drop this data to avoid later problems (defaults to TRUE)
#' @param numcores number of cores to use in parallel loop. To check number of available cores: parallel::detectCores().
#' If you receive an error regarding the speed you are making calls, reduce this number
#' @param bypassNumCallCheck set to TRUE to avoid prompting the user to confirm that they want to begin making API calls
#' @param returnSpatialData returns the data as a SpatialPixels object. Can be convered to raster with the command raster::stack
#' NOTE: if multiple days worth of data is returned, it is necessary to subset to specific day for working with
#' as spatial data (sp package: optional)
#' @param verbose Set to TRUE tp print messages to console about state of parallization call. Typically only visible if run from console and not GUI
#' @param maxTryCount maximum number of times a call is repeated if the the API returns an error. Random pause between each call
#' @param keyToUse aWhere API key to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param secretToUse aWhere API secret to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param tokenToUse: aWhere API token to use. For advanced use only. Most users will not need to use this parameter. Note that if you specify
#' your own token there is no functionality in this function for requesting a new token if the one originally used expires while
#' requesting data. Use at your own risk (optional)
#' @param apiAddressToUse Address of aWhere API to use. For advanced use only. Most users will not need to use this parameter (optional)
#'
#' @import httr
#' @import data.table
#' @import lubridate
#' @import jsonlite
#' @import raster
#' @import foreach
#' @import doParallel
#' @import rgeos
#'
#' @return data.frame of requested data for dates requested
#'
#' @examples
#' \dontrun{weather_norms_area(polygon = raster::getData('GADM', country = "Gambia", level = 0, download = T)
#' ,monthday_start = '02-01'
#' ,monthday_end = '03-10'
#' ,year_start = 2008
#' ,year_end = 2015
#' ,exclude_years = c(2010,2011)
#' ,numcores = 2)}
#' @export
weather_norms_area <- function(polygon
,monthday_start
,monthday_end
,year_start
,year_end
,propertiesToInclude = ''
,exclude_years = NULL
,includeFeb29thData = TRUE
,numcores = 2
,bypassNumCallCheck = FALSE
,returnSpatialData = FALSE
,verbose = TRUE
,maxTryCount = 3
,keyToUse = awhereEnv75247$uid
,secretToUse = awhereEnv75247$secret
,tokenToUse = awhereEnv75247$token
,apiAddressToUse = awhereEnv75247$apiAddress) {
#Checking Input Parameters
checkCredentials(keyToUse,secretToUse,tokenToUse)
checkNormsStartEndDates(monthday_start,monthday_end)
checkNormsYearsToRequest(year_start,year_end,monthday_start,monthday_end,exclude_years)
checkPropertiesEndpoint('weather_norms',propertiesToInclude)
##############################################################################
if (tokenToUse == awhereEnv75247$token) {
useTokenFromEnv <- TRUE
} else {
useTokenFromEnv <- FALSE
}
if (!(all(class(polygon) %in% c('data.frame','data.table')))) {
if (verbose == TRUE) {
cat(paste0('Creating aWhere Raster Grid within Polygon\n'))
}
grid <- create_awhere_grid(polygon)
} else {
if (!(all(colnames(polygon) %in% c('lat','lon')) & length(colnames(polygon)) == 2)) {
stop('Data.Frame of Lat/Lon coordinates improperly specified, please correct')
}
grid <- data.table::as.data.table(polygon)
grid[,c('gridx'
,'gridy') := list(getGridX(longitude = lon)
,getGridY(latitude = lat))]
}
verify_api_calls(grid,bypassNumCallCheck)
if (verbose == TRUE) {
cat(paste0('Requesting data using parallal API calls\n'))
}
grid <- split(grid, seq(1,nrow(grid),1))
if (numcores > 1) {
doParallel::registerDoParallel(cores=numcores)
`%loopToUse%` <- `%dopar%`
} else {
`%loopToUse%` <- `%do%`
}
if (length(grid) > 1000) {
howOftenPrintVerbose <- 100
} else if (length(grid) > 500) {
howOftenPrintVerbose <- 50
} else if (length(grid) > 100) {
howOftenPrintVerbose <- 25
} else {
howOftenPrintVerbose <- 10
}
norms <- foreach::foreach(j=c(1:length(grid))
,.packages = c("aWhereAPI")
,.errorhandling = 'pass') %loopToUse% {
if (verbose == TRUE & (j == 1 | (j %% howOftenPrintVerbose) == 0)) {
cat(paste0(' Currently requesting data for location ',j,' of ',length(grid),'\n'))
}
tryCount <- 1
while (tryCount < maxTryCount) {
#this works because if no error occurs the loop will return the data
#given by the API. If an error is received it will increment the
#tryCount timer and repear
tryCount <-
tryCatch({
if (useTokenFromEnv == TRUE) {
tokenToUse = awhereEnv75247$token
}
t <-
weather_norms_latlng(latitude = grid[[j]]$lat
,longitude = grid[[j]]$lon
,monthday_start = monthday_start
,monthday_end = monthday_end
,year_start = year_start
,year_end = year_end
,propertiesToInclude = propertiesToInclude
,exclude_years = exclude_years
,includeFeb29thData = includeFeb29thData
,keyToUse = keyToUse
,secretToUse = secretToUse
,tokenToUse = tokenToUse
,apiAddressToUse = apiAddressToUse)
currentNames <- colnames(t)
t$gridy <- grid[[j]]$gridy
t$gridx <- grid[[j]]$gridx
data.table::setcolorder(t, c(currentNames[c(1:2)], "gridy", "gridx", currentNames[c(3:length(currentNames))]))
return(t)
}, error = function(e) {
cat(paste0(' Error received from API on location ',j,': Try ',tryCount,'\n'))
Sys.sleep(runif(n = 1
,min = 10
,max = 30))
tryCount <- tryCount + 1
tryCount
})
if (tryCount >= maxTryCount) {
cat(paste0(' NO DATA WAS ABLE TO RETRIEVED FROM API FOR LOCATION ',j,'\n'))
return(simpleError(message = 'Consecutive Errors from API\n'))
}
}
}
grid <- data.table::rbindlist(grid)
indexToRemove <- c()
for (x in 1:length(norms)) {
if (any(class(norms[[x]]) == 'simpleError')) {
indexToRemove <- c(indexToRemove,x)
}
}
if (length(indexToRemove) > 0) {
warning(paste0('The following locations returned errors and have been removed from the output. Please investigate by running manually:\n'
,paste0(grid[indexToRemove,paste0('(',lat,', ',lon,')')],collapse = ', ')
,'\n'))
grid <- grid[!indexToRemove]
norms[indexToRemove] <- NULL
}
norms <- data.table::rbindlist(norms,use.names = TRUE,fill = TRUE)
if (returnSpatialData == TRUE) {
sp::coordinates(norms) <- ~longitude + latitude
sp::proj4string(norms) <- sp::CRS("+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0")
sp::gridded(norms) <- TRUE
return(norms)
}
return(as.data.frame(norms))
}
aWhereAPI/aWhere-R-Library documentation built on Nov. 5, 2021, 3:35 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions weather_norms_area weather_norms_latlng weather_norms_fields
Documented in weather_norms_area weather_norms_fields weather_norms_latlng
#' @title weather_norms_fields
#'
#' @description
#' \code{weather_norms_fields} pulls long term norm weather data from aWhere's API based on field id
#'
#' @details
#' This function allows you to calculate the averages for weather attributes
#' across any range of years for which data are available. The data pulled includes
#' meanTemp, maxTemp, minTemp, precipitation average, solar radiation average,
#' minHumidity, maxHumidity, maxWind and averageWind, along with the standard deviations
#' for these variables. The data pulled is for the field id identified.
#'
#' The data returned in this function
#' allow you to compare this year or previous years to the long-term normals, calculated as
#' the average of those weather conditions on that day in that location over the years specified.
#'
#' Note about dates: The system does not adjust for any difference in dates between the location of the user
#' and where data is being requested from. It is the responsibility of the user to ensure a valid
#' date range is specified given any differences in timezone. These differences can have implications
#' for whether a given date should be requested from the daily_observed functions or the forecast functions
#'
#' @references https://docs.awhere.com/knowledge-base-docs/historical-weather-norms/
#'
#' @param field_id the field_id associated with the location for which you want to pull data.
#' Field IDs are created using the create_field function. (string)
#' @param monthday_start character string of the first month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the start of your date range. e.g. '07-01' (July 1) (required)
#' @param monthday_end character string of the last month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the end of your date range. e.g. '07-01' (July 1) (required)
#' @param year_start character string of the starting year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2008 (required)
#' @param year_end character string of the last year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2015 (required)
#' @param propertiesToInclude character vector of properties to retrieve from API.
#' Valid values are meanTemp, maxTemp, minTemp, precipitation,
#' solar, maxHumidity, minHumidity, dailyMaxWind (optional)
#' @param exclude_year Year or years which you'd like to exclude from
#' your range of years on which to calculate norms. To exclude
#' multiple years, provide a vector of years. You must include
#' at least three years of data with which to calculate the norms. (numeric, optional)
#' @param includeFeb29thData Whether to keep data from Feb 29th on leap years. Because weather/agronomics
#' summary statistics are calculated via the calendar date and 3 years are required
#' to generate a value, data from this date is more likely to be NA. ALlows user
#' to drop this data to avoid later problems (defaults to TRUE)
#' @param keyToUse aWhere API key to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param secretToUse aWhere API secret to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param tokenToUse aWhere API token to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param apiAddressToUse Address of aWhere API to use. For advanced use only. Most users will not need to use this parameter (optional)
#'
#' @import httr
#' @import data.table
#' @import lubridate
#' @import jsonlite
#'
#' @return dataframe of requested data for dates requested
#'
#' @examples
#' \dontrun{weather_norms_fields(field_id = "field_test"
#' ,monthday_start = "06-01"
#' ,monthday_end = "09-01"
#' ,year_start = 2006
#' ,year_end = 2015)}
#' @export
#'
weather_norms_fields <- function(field_id
,monthday_start
,monthday_end
,year_start
,year_end
,propertiesToInclude = ''
,exclude_years = NULL
,includeFeb29thData = TRUE
,keyToUse = awhereEnv75247$uid
,secretToUse = awhereEnv75247$secret
,tokenToUse = awhereEnv75247$token
,apiAddressToUse = awhereEnv75247$apiAddress) {
#Checking Input Parameters
checkCredentials(keyToUse,secretToUse,tokenToUse)
checkValidField(field_id,keyToUse,secretToUse,tokenToUse)
checkNormsStartEndDates(monthday_start,monthday_end)
checkNormsYearsToRequest(year_start,year_end,monthday_start,monthday_end,exclude_years)
checkPropertiesEndpoint('weather_norms',propertiesToInclude)
# Create Logic of API Request
numObsReturned <- 120
calculateAPIRequests <- TRUE
continueRequestingData <- TRUE
yearsToInclude <- setdiff(seq(year_start,year_end,1),exclude_years)
dataList <- list()
# loop through, making requests in chunks of size numObsReturned
while (continueRequestingData == TRUE | calculateAPIRequests == TRUE) {
#If this clause is triggered the progression of API calls will be
#calculated. After each API call the return will be checked for an error
#indicating that the request was too large. If that occurs this loop will
#be reenentered to calculate using the smaller return size
############################################################################
if (calculateAPIRequests == TRUE) {
calculateAPIRequests <- FALSE
#We need to consider whether a year year is present to determine if Fe
#29th data will be returned
includesLeapYear <- any(is.leapyear(yearsToInclude))
if (includesLeapYear == TRUE) {
yearPrefix <- paste0(yearsToInclude[is.leapyear(yearsToInclude)][1],'-')
} else {
yearPrefix <- paste0(yearsToInclude[1],'-')
}
day_start <- ymd(paste0(yearPrefix
,monthday_start))
day_end <- ymd(paste0(yearPrefix
,monthday_end))
temp <- plan_APICalls(day_start
,day_end
,numObsReturned
,includesLeapYear)
allDates <- temp[[1]]
loops <- temp[[2]]
#remove the years
allDates <- gsub(pattern = '20\\d\\d-',replacement = '',x = allDates)
}
#This for loop will make the API requests as calculated from above
############################################################################
for (i in 1:loops) {
starting = numObsReturned*(i-1)+1
ending = numObsReturned*i
monthday_start_toUse <- allDates[starting]
monthday_end_toUse <- allDates[ending]
if(is.na(monthday_end_toUse)) {
tempDates <- allDates[c(starting:length(allDates))]
monthday_start_toUse <- tempDates[1]
monthday_end_toUse <- tempDates[length(tempDates)]
}
##############################################################################
# Create query
urlAddress <- paste0(apiAddressToUse, "/weather")
strBeg <- paste0('/fields')
strCoord <- paste0('/',field_id)
strType <- paste0('/norms')
strMonthsDays <- paste0('/',monthday_start_toUse,',',monthday_end_toUse)
limitString <- paste0('?limit=',numObsReturned)
if (length(exclude_years) != 0) {
strexclude_years <- paste0('&excludeYears=',toString(exclude_years))
} else {
strexclude_years <- ''
}
if (propertiesToInclude[1] != '') {
propertiesString <- paste0('&properties=',paste0(propertiesToInclude,collapse = ','))
} else {
propertiesString <- ''
}
strYearsType <- paste0('/years')
strYears <- paste0('/',year_start,',',year_end)
url <- URLencode(paste0(urlAddress
,strBeg
,strCoord
,strType
,strMonthsDays
,strYearsType
,strYears
,limitString
,strexclude_years
,propertiesString))
doWeatherGet <- TRUE
tryCount <- 0
while (doWeatherGet == TRUE) {
tryCount <- tryCount + 1
postbody = ''
request <- httr::GET(url, body = postbody, httr::content_type('application/json'),
httr::add_headers(Authorization =paste0("Bearer ", tokenToUse)))
a <- suppressMessages(httr::content(request, as = "text"))
temp <- check_JSON(a
,request
,tryCount
,keyToUse
,secretToUse
,tokenToUse)
doWeatherGet <- temp[[1]]
#if the token was updated, this will cause it to be used through function
tokenToUse <- temp[[3]]
#The temp[[2]] will only not be NA when the limit param is too large.
if(!is.na(temp[[2]] == TRUE)) {
numObsReturned <- temp[[2]]
goodReturn <- FALSE
break
} else {
goodReturn <- TRUE
}
rm(temp)
}
if (goodReturn == TRUE) {
#The JSONLITE Serializer properly handles the JSON conversion
x <- jsonlite::fromJSON(a,flatten = TRUE)
data <- data.table::as.data.table(x[[1]])
data <- removeUnnecessaryColumns(data)
dataList[[length(dataList) + 1]] <- data
} else {
#This will break out of the current loop of making API requests so that
#the logic of the API requests can be recalculated
calculateAPIRequests <- TRUE
break
}
}
continueRequestingData <- FALSE
}
data <- unique(rbindlist(dataList))
#Get rid of leap yearData
if (includeFeb29thData == FALSE) {
data <- data[day != '02-29',]
}
currentNames <- data.table::copy(colnames(data))
data[,field_id := field_id]
data.table::setcolorder(data,c('field_id',currentNames))
checkDataReturn_norms(data,monthday_start,monthday_end,year_start,year_end,exclude_years,includeFeb29thData)
return(as.data.frame(data))
}
#' @title weather_norms_latlng
#'
#' @description
#' \code{weather_norms_latlng} pulls long term norm weather data from aWhere's API based on latitude & longitude
#'
#' @details
#' This function allows you to calculate the averages for weather attributes
#' across any range of years for which data are available. The data pulled includes
#' meanTemp, maxTemp, minTemp, precipitation average, solar radiation average,
#' minHumidity, maxHumidity, maxWind and averageWind, along with the standard deviations
#' for these variables. The data pulled is for the latitude & longitude identified.
#'
#' The data returned in this function
#' allow you to compare this year or previous years to the long-term normals, calculated as
#' the average of those weather conditions on that day in that location over the years specified.
#'
#' Note about dates: The system does not adjust for any difference in dates between the location of the user
#' and where data is being requested from. It is the responsibility of the user to ensure a valid
#' date range is specified given any differences in timezone. These differences can have implications
#' for whether a given date should be requested from the daily_observed functions or the forecast functions
#'
#' @references https://docs.awhere.com/knowledge-base-docs/historical-weather-norms-by-geolocation/
#'
#' @param latitude the latitude of the requested location (double, required)
#' @param longitude the longitude of the requested locations (double, required)
#' @param monthday_start character string of the first month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the start of your date range. e.g. '07-01' (July 1) (required)
#' @param monthday_end character string of the last month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the end of your date range. e.g. '07-01' (July 1) (required)
#' @param year_start character string of the starting year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2008 (required)
#' @param year_end character string of the last year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2015 (required)
#' @param propertiesToInclude character vector of properties to retrieve from API.
#' Valid values are meanTemp, maxTemp, minTemp, precipitation,
#' solar, maxHumidity, minHumidity, dailyMaxWind (optional)
#' @param exclude_year Year or years which you'd like to exclude from
#' your range of years on which to calculate norms. To exclude
#' multiple years, provide a vector of years. You must include
#' at least three years of data with which to calculate the norms. (numeric, optional)
#' @param includeFeb29thData Whether to keep data from Feb 29th on leap years. Because weather/agronomics
#' summary statistics are calculated via the calendar date and 3 years are required
#' to generate a value, data from this date is more likely to be NA. ALlows user
#' to drop this data to avoid later problems (defaults to TRUE)
#' @param keyToUse aWhere API key to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param secretToUse aWhere API secret to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param tokenToUse aWhere API token to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param apiAddressToUse: Address of aWhere API to use. For advanced use only. Most users will not need to use this parameter (optional)
#'
#' @import httr
#' @import data.table
#' @import lubridate
#' @import jsonlite
#'
#' @return data.frame of requested data for dates requested
#'
#' @examples
#' \dontrun{weather_norms_latlng(latitude = 39.8282
#' ,longitude = -98.5795
#' ,monthday_start = '02-01'
#' ,monthday_end = '03-10'
#' ,year_start = 2008
#' ,year_end = 2015
#' ,exclude_years = c(2010,2011))}
#' @export
weather_norms_latlng <- function(latitude
,longitude
,monthday_start
,monthday_end
,year_start
,year_end
,propertiesToInclude = ''
,exclude_years = NULL
,includeFeb29thData = TRUE
,keyToUse = awhereEnv75247$uid
,secretToUse = awhereEnv75247$secret
,tokenToUse = awhereEnv75247$token
,apiAddressToUse = awhereEnv75247$apiAddress) {
#Checking Input Parameters
checkCredentials(keyToUse,secretToUse,tokenToUse)
checkValidLatLong(latitude,longitude)
checkNormsStartEndDates(monthday_start,monthday_end)
checkNormsYearsToRequest(year_start,year_end,monthday_start,monthday_end,exclude_years)
checkPropertiesEndpoint('weather_norms',propertiesToInclude)
# Create Logic of API Request
numObsReturned <- 120
calculateAPIRequests <- TRUE
continueRequestingData <- TRUE
yearsToInclude <- setdiff(seq(year_start,year_end,1),exclude_years)
dataList <- list()
# loop through, making requests in chunks of size numObsReturned
while (continueRequestingData == TRUE | calculateAPIRequests == TRUE) {
#If this clause is triggered the progression of API calls will be
#calculated. After each API call the return will be checked for an error
#indicating that the request was too large. If that occurs this loop will
#be reenentered to calculate using the smaller return size
############################################################################
if (calculateAPIRequests == TRUE) {
calculateAPIRequests <- FALSE
#We need to consider whether a year year is present to determine if Fe
#29th data will be returned
includesLeapYear <- any(is.leapyear(yearsToInclude))
if (includesLeapYear == TRUE) {
yearPrefix <- paste0(yearsToInclude[is.leapyear(yearsToInclude)][1],'-')
} else {
yearPrefix <- paste0(yearsToInclude[1],'-')
}
day_start <- ymd(paste0(yearPrefix
,monthday_start))
day_end <- ymd(paste0(yearPrefix
,monthday_end))
temp <- plan_APICalls(day_start
,day_end
,numObsReturned
,includesLeapYear)
allDates <- temp[[1]]
loops <- temp[[2]]
#remove the years
allDates <- gsub(pattern = '20\\d\\d-',replacement = '',x = allDates)
}
#This for loop will make the API requests as calculated from above
############################################################################
for (i in 1:loops) {
starting = numObsReturned*(i-1)+1
ending = numObsReturned*i
monthday_start_toUse <- allDates[starting]
monthday_end_toUse <- allDates[ending]
if(is.na(monthday_end_toUse)) {
tempDates <- allDates[c(starting:length(allDates))]
monthday_start_toUse <- tempDates[1]
monthday_end_toUse <- tempDates[length(tempDates)]
}
##############################################################################
# Create query
urlAddress <- paste0(apiAddressToUse, "/weather")
strBeg <- paste0('/locations')
strCoord <- paste0('/',latitude,',',longitude)
strType <- paste0('/norms')
strMonthsDays <- paste0('/',monthday_start_toUse,',',monthday_end_toUse)
limitString <- paste0('?limit=',numObsReturned)
if (length(exclude_years) != 0) {
strexclude_years <- paste0('&excludeYears=',toString(exclude_years))
} else {
strexclude_years <- ''
}
if (propertiesToInclude[1] != '') {
propertiesString <- paste0('&properties=',paste0(propertiesToInclude,collapse = ','))
} else {
propertiesString <- ''
}
strYearsType <- paste0('/years')
strYears <- paste0('/',year_start,',',year_end)
url <- URLencode(paste0(urlAddress
,strBeg
,strCoord
,strType
,strMonthsDays
,strYearsType
,strYears
,limitString
,strexclude_years
,propertiesString))
doWeatherGet <- TRUE
tryCount <- 0
while (doWeatherGet == TRUE) {
tryCount <- tryCount + 1
postbody = ''
request <- httr::GET(url, body = postbody, httr::content_type('application/json'),
httr::add_headers(Authorization =paste0("Bearer ", tokenToUse)))
a <- suppressMessages(httr::content(request, as = "text"))
temp <- check_JSON(a
,request
,tryCount
,keyToUse
,secretToUse
,tokenToUse)
doWeatherGet <- temp[[1]]
#if the token was updated, this will cause it to be used through function
tokenToUse <- temp[[3]]
#The temp[[2]] will only not be NA when the limit param is too large.
if(!is.na(temp[[2]] == TRUE)) {
numObsReturned <- temp[[2]]
goodReturn <- FALSE
break
} else {
goodReturn <- TRUE
}
rm(temp)
}
if (goodReturn == TRUE) {
#The JSONLITE Serializer properly handles the JSON conversion
x <- jsonlite::fromJSON(a,flatten = TRUE)
data <- data.table::as.data.table(x[[1]])
data <- removeUnnecessaryColumns(data)
dataList[[length(dataList) + 1]] <- data
} else {
#This will break out of the current loop of making API requests so that
#the logic of the API requests can be recalculated
calculateAPIRequests <- TRUE
break
}
}
continueRequestingData <- FALSE
}
data <- unique(rbindlist(dataList))
#Get rid of leap yearData
if (includeFeb29thData == FALSE) {
data <- data[day != '02-29',]
}
currentNames <- data.table::copy(colnames(data))
data[,latitude := latitude]
data[,longitude := longitude]
data.table::setcolorder(data,c('latitude','longitude',currentNames))
checkDataReturn_norms(data,monthday_start,monthday_end,year_start,year_end,exclude_years,includeFeb29thData)
return(as.data.frame(data))
}
#' @title weather_norms_area
#'
#' @description
#' \code{weather_norms_area} pulls long term norm weather data from aWhere's API based on a data.frame of lat/lon, polygon or extentt
#'
#' @details
#' This function allows you to calculate the averages for weather attributes
#' across any range of years for which data are available. The data pulled includes
#' meanTemp, maxTemp, minTemp, precipitation average, solar radiation average,
#' minHumidity, maxHumidity, maxWind and averageWind, along with the standard deviations
#' for these variables. The polygon should be either a SpatialPolygons object or
#' a well-known text character string or an extent.
#'
#' The data returned in this function
#' allow you to compare this year or previous years to the long-term normals, calculated as
#' the average of those weather conditions on that day in that location over the years specified.
#'
#' Note about dates: The system does not adjust for any difference in dates between the location of the user
#' and where data is being requested from. It is the responsibility of the user to ensure a valid
#' date range is specified given any differences in timezone. These differences can have implications
#' for whether a given date should be requested from the daily_observed functions or the forecast functions.
#' Furthermore, because this function can take as input locations that may be in different timezones, it is
#' the responsibility of the user to either ensure that the date range specified is valid for all relevant
#' locations or to break the query into pieces.
#'
#' @references https://docs.awhere.com/knowledge-base-docs/historical-weather-norms-by-geolocation/
#'
#' @param polygon either a data.frame with column names lat/lon, SpatialPolygons object,
#' well-known text string, or extent from raster package. If the object contains
#' multiple polygons, the union of them is used. Information from each individal
#' polygon can be retrieved by returning spatial data and using
#' the over function from the sp package
#' @param monthday_start character string of the first month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the start of your date range. e.g. '07-01' (July 1) (required)
#' @param monthday_end character string of the last month and day for which you want to retrieve data,
#' in the form: MM-DD. This is the end of your date range. e.g. '07-01' (July 1) (required)
#' @param year_start character string of the starting year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2008 (required)
#' @param year_end character string of the last year (inclusive) of the range of years for which
#' you're calculating norms, in the form YYYY. e.g., 2015 (required)
#' @param propertiesToInclude character vector of properties to retrieve from API. Valid values are meanTemp, maxTemp, minTemp, precipitation, solar, maxHumidity, minHumidity, dailyMaxWind (optional)
#' @param exclude_year Year or years which you'd like to exclude from
#' your range of years on which to calculate norms. To exclude
#' multiple years, provide a vector of years. You must include
#' at least three years of data with which to calculate the norms. (numeric, optional)
#' @param includeFeb29thData Whether to keep data from Feb 29th on leap years. Because weather/agronomics
#' summary statistics are calculated via the calendar date and 3 years are required
#' to generate a value, data from this date is more likely to be NA. ALlows user
#' to drop this data to avoid later problems (defaults to TRUE)
#' @param numcores number of cores to use in parallel loop. To check number of available cores: parallel::detectCores().
#' If you receive an error regarding the speed you are making calls, reduce this number
#' @param bypassNumCallCheck set to TRUE to avoid prompting the user to confirm that they want to begin making API calls
#' @param returnSpatialData returns the data as a SpatialPixels object. Can be convered to raster with the command raster::stack
#' NOTE: if multiple days worth of data is returned, it is necessary to subset to specific day for working with
#' as spatial data (sp package: optional)
#' @param verbose Set to TRUE tp print messages to console about state of parallization call. Typically only visible if run from console and not GUI
#' @param maxTryCount maximum number of times a call is repeated if the the API returns an error. Random pause between each call
#' @param keyToUse aWhere API key to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param secretToUse aWhere API secret to use. For advanced use only. Most users will not need to use this parameter (optional)
#' @param tokenToUse: aWhere API token to use. For advanced use only. Most users will not need to use this parameter. Note that if you specify
#' your own token there is no functionality in this function for requesting a new token if the one originally used expires while
#' requesting data. Use at your own risk (optional)
#' @param apiAddressToUse Address of aWhere API to use. For advanced use only. Most users will not need to use this parameter (optional)
#'
#' @import httr
#' @import data.table
#' @import lubridate
#' @import jsonlite
#' @import raster
#' @import foreach
#' @import doParallel
#' @import rgeos
#'
#' @return data.frame of requested data for dates requested
#'
#' @examples
#' \dontrun{weather_norms_area(polygon = raster::getData('GADM', country = "Gambia", level = 0, download = T)
#' ,monthday_start = '02-01'
#' ,monthday_end = '03-10'
#' ,year_start = 2008
#' ,year_end = 2015
#' ,exclude_years = c(2010,2011)
#' ,numcores = 2)}
#' @export
weather_norms_area <- function(polygon
,monthday_start
,monthday_end
,year_start
,year_end
,propertiesToInclude = ''
,exclude_years = NULL
,includeFeb29thData = TRUE
,numcores = 2
,bypassNumCallCheck = FALSE
,returnSpatialData = FALSE
,verbose = TRUE
,maxTryCount = 3
,keyToUse = awhereEnv75247$uid
,secretToUse = awhereEnv75247$secret
,tokenToUse = awhereEnv75247$token
,apiAddressToUse = awhereEnv75247$apiAddress) {
#Checking Input Parameters
checkCredentials(keyToUse,secretToUse,tokenToUse)
checkNormsStartEndDates(monthday_start,monthday_end)
checkNormsYearsToRequest(year_start,year_end,monthday_start,monthday_end,exclude_years)
checkPropertiesEndpoint('weather_norms',propertiesToInclude)
##############################################################################
if (tokenToUse == awhereEnv75247$token) {
useTokenFromEnv <- TRUE
} else {
useTokenFromEnv <- FALSE
}
if (!(all(class(polygon) %in% c('data.frame','data.table')))) {
if (verbose == TRUE) {
cat(paste0('Creating aWhere Raster Grid within Polygon\n'))
}
grid <- create_awhere_grid(polygon)
} else {
if (!(all(colnames(polygon) %in% c('lat','lon')) & length(colnames(polygon)) == 2)) {
stop('Data.Frame of Lat/Lon coordinates improperly specified, please correct')
}
grid <- data.table::as.data.table(polygon)
grid[,c('gridx'
,'gridy') := list(getGridX(longitude = lon)
,getGridY(latitude = lat))]
}
verify_api_calls(grid,bypassNumCallCheck)
if (verbose == TRUE) {
cat(paste0('Requesting data using parallal API calls\n'))
}
grid <- split(grid, seq(1,nrow(grid),1))
if (numcores > 1) {
doParallel::registerDoParallel(cores=numcores)
`%loopToUse%` <- `%dopar%`
} else {
`%loopToUse%` <- `%do%`
}
if (length(grid) > 1000) {
howOftenPrintVerbose <- 100
} else if (length(grid) > 500) {
howOftenPrintVerbose <- 50
} else if (length(grid) > 100) {
howOftenPrintVerbose <- 25
} else {
howOftenPrintVerbose <- 10
}
norms <- foreach::foreach(j=c(1:length(grid))
,.packages = c("aWhereAPI")
,.errorhandling = 'pass') %loopToUse% {
if (verbose == TRUE & (j == 1 | (j %% howOftenPrintVerbose) == 0)) {
cat(paste0(' Currently requesting data for location ',j,' of ',length(grid),'\n'))
}
tryCount <- 1
while (tryCount < maxTryCount) {
#this works because if no error occurs the loop will return the data
#given by the API. If an error is received it will increment the
#tryCount timer and repear
tryCount <-
tryCatch({
if (useTokenFromEnv == TRUE) {
tokenToUse = awhereEnv75247$token
}
t <-
weather_norms_latlng(latitude = grid[[j]]$lat
,longitude = grid[[j]]$lon
,monthday_start = monthday_start
,monthday_end = monthday_end
,year_start = year_start
,year_end = year_end
,propertiesToInclude = propertiesToInclude
,exclude_years = exclude_years
,includeFeb29thData = includeFeb29thData
,keyToUse = keyToUse
,secretToUse = secretToUse
,tokenToUse = tokenToUse
,apiAddressToUse = apiAddressToUse)
currentNames <- colnames(t)
t$gridy <- grid[[j]]$gridy
t$gridx <- grid[[j]]$gridx
data.table::setcolorder(t, c(currentNames[c(1:2)], "gridy", "gridx", currentNames[c(3:length(currentNames))]))
return(t)
}, error = function(e) {
cat(paste0(' Error received from API on location ',j,': Try ',tryCount,'\n'))
Sys.sleep(runif(n = 1
,min = 10
,max = 30))
tryCount <- tryCount + 1
tryCount
})
if (tryCount >= maxTryCount) {
cat(paste0(' NO DATA WAS ABLE TO RETRIEVED FROM API FOR LOCATION ',j,'\n'))
return(simpleError(message = 'Consecutive Errors from API\n'))
}
}
}
grid <- data.table::rbindlist(grid)
indexToRemove <- c()
for (x in 1:length(norms)) {
if (any(class(norms[[x]]) == 'simpleError')) {
indexToRemove <- c(indexToRemove,x)
}
}
if (length(indexToRemove) > 0) {
warning(paste0('The following locations returned errors and have been removed from the output. Please investigate by running manually:\n'
,paste0(grid[indexToRemove,paste0('(',lat,', ',lon,')')],collapse = ', ')
,'\n'))
grid <- grid[!indexToRemove]
norms[indexToRemove] <- NULL
}
norms <- data.table::rbindlist(norms,use.names = TRUE,fill = TRUE)
if (returnSpatialData == TRUE) {
sp::coordinates(norms) <- ~longitude + latitude
sp::proj4string(norms) <- sp::CRS("+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0")
sp::gridded(norms) <- TRUE
return(norms)
}
return(as.data.frame(norms))
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.