importNGWMN_wml2.R 5.25 KB
Newer Older
David Watkins's avatar
David Watkins committed
1
#' Function to return data from the National Ground Water Monitoring Network waterML2 format
2
3
4
#'
#' This function accepts a url parameter for a WaterML2 getObservation. This function is still under development,
#' but the general functionality is correct.
David Watkins's avatar
David Watkins committed
5
#' 
6
7
8
#' @param input character or raw, containing the url for the retrieval or a path to the data file, or raw XML.
#' @param asDateTime logical, if \code{TRUE} returns date and time as POSIXct, if \code{FALSE}, character
#' @param tz character to set timezone attribute of datetime. Default is an empty quote, which converts the 
David Watkins's avatar
David Watkins committed
9
#' datetimes to UTC (properly accounting for daylight savings times based on the data's provided time zone offset).
10
11
#' Possible values to provide are "America/New_York","America/Chicago", "America/Denver","America/Los_Angeles",
#' "America/Anchorage","America/Honolulu","America/Jamaica","America/Managua","America/Phoenix", and "America/Metlakatla"
David Watkins's avatar
David Watkins committed
12
#' @return mergedDF a data frame source, time, value, uom, uomTitle, comment, gmlID
13
14
15
16
17
18
19
#' @export
#' @importFrom xml2 read_xml
#' @importFrom xml2 xml_find_all
#' @importFrom xml2 xml_text
#' @importFrom xml2 xml_attr
#' @importFrom lubridate parse_date_time
#' @examples
David Watkins's avatar
David Watkins committed
20
21
22
#' \dontrun{
#' url <- "http://cida.usgs.gov/ngwmn_cache/sos?request=GetObservation&service=SOS&version=2.0.0&observedProperty=urn:ogc:def:property:OGC:GroundWaterLevel&responseFormat=text/xml&featureOfInterest=VW_GWDP_GEOSERVER.USGS.403836085374401"
#' data <- importNGWMN_wml2(url)
David Watkins's avatar
David Watkins committed
23
24
25
#' 
#' url <- "http://cida.usgs.gov/ngwmn_cache/sos?request=GetObservation&service=SOS&version=2.0.0&observedProperty=urn:ogc:def:property:OGC:GroundWaterLevel&responseFormat=text/xml&featureOfInterest=VW_GWDP_GEOSERVER.USGS.474011117072901"
#' data <- importNGWMN_wml2(url)
David Watkins's avatar
David Watkins committed
26
#' }
David Watkins's avatar
David Watkins committed
27
28
#' 
importNGWMN_wml2 <- function(input, asDateTime=FALSE, tz=""){
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
  if(tz != ""){
    tz <- match.arg(tz, c("America/New_York","America/Chicago",
                          "America/Denver","America/Los_Angeles",
                          "America/Anchorage","America/Honolulu",
                          "America/Jamaica","America/Managua",
                          "America/Phoenix","America/Metlakatla"))
  }else{tz = "UTC"}
  
  raw <- FALSE
  if(class(input) == "character" && file.exists(input)){
    returnedDoc <- read_xml(input)
  }else if(class(input) == 'raw'){
    returnedDoc <- read_xml(input)
    raw <- TRUE
  } else {
    returnedDoc <- xml_root(getWebServiceData(input, encoding='gzip'))
  }
  
David Watkins's avatar
David Watkins committed
47
48
49
50
51
52
53
54
55
56
57
  response <- xml_name(returnedDoc)
  if(response == "GetObservationResponse"){
    
    timeSeries <- xml_find_all(returnedDoc, "//wml2:MeasurementTimeseries") #each parameter/site combo
    
    if(0 == length(timeSeries)){
      df <- data.frame()
      if(!raw){
        attr(df, "url") <- input
      }
      return(df)
58
59
    }
    
David Watkins's avatar
David Watkins committed
60
    mergedDF <- NULL
61
    
David Watkins's avatar
David Watkins committed
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
    for(t in timeSeries){
      gmlID <- xml_attr(t,"id")
      TVP <- xml_find_all(t, ".//wml2:MeasurementTVP")#time-value pairs
      time <- xml_text(xml_find_all(TVP,".//wml2:time"))
      if(asDateTime){
        time <- parse_date_time(time, c("%Y","%Y-%m-%d","%Y-%m-%dT%H:%M","%Y-%m-%dT%H:%M:%S",
                                        "%Y-%m-%dT%H:%M:%OS","%Y-%m-%dT%H:%M:%OS%z"), exact = TRUE)
        #^^setting tz in as.POSIXct just sets the attribute, does not convert the time!
        attr(time, 'tzone') <- tz 
      }
      valueNodes <- xml_find_all(TVP,".//wml2:value")
      values <- as.numeric(xml_text(valueNodes))
      nVals <- length(values)
      gmlID <- rep(gmlID, nVals)
      
      uom <- removeBlanks(xml_attr(valueNodes, "uom", default = NA))
      source <- xml_attr(xml_find_all(TVP, ".//wml2:source"), "title")
      comment <- removeBlanks(xml_text(xml_find_all(TVP, ".//wml2:comment")))
      
      df <- cbind.data.frame(source, time, value=values, uom, comment, gmlID,
                             stringsAsFactors=FALSE)
      if (is.null(mergedDF)){
        mergedDF <- df
      } else {
        similarNames <- intersect(colnames(mergedDF), colnames(df))
        mergedDF <- full_join(mergedDF, df, by=similarNames)
      }
89
    }
David Watkins's avatar
David Watkins committed
90
91
92
93
94
95
96
    
    if(!raw){
      url <- input
      attr(mergedDF, "url") <- url
    }
    attr(mergedDF, "gml:identifier") <- xml_text(xml_find_all(returnedDoc, ".//gml:identifier")) 
    attr(mergedDF, "generationDate") <- xml_text(xml_find_all(returnedDoc, ".//wml2:generationDate")) 
David Watkins's avatar
David Watkins committed
97
    mergedDF$value[mergedDF$value == -999999.0] <- NA
David Watkins's avatar
David Watkins committed
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
    
  }else if(response == "GetFeatureOfInterestResponse"){
    site <- xml_text(xml_find_all(returnedDoc,".//gml:identifier"))
    site <- substring(site, 8)
    
    #bandaid to work with only single site calls
    #TODO: need better solution when bbox is added
    siteDesc <- xml_text(xml_find_all(returnedDoc, ".//gml:description"))
    if(length(siteDesc) == 0){
      siteDesc <- NA
    }
    
    siteLocs <- strsplit(xml_text(xml_find_all(returnedDoc, ".//gml:pos")), " ")
    siteLocs <- data.frame(lat=siteLocs[[1]][1], lon=siteLocs[[1]][2], stringsAsFactors = FALSE)
    mergedDF <- cbind.data.frame(site, description = siteDesc, siteLocs, stringsAsFactors = FALSE) 
  }
David Watkins's avatar
David Watkins committed
114
115
  else{
    stop("Unrecognized response from the web service")
116
117
118
119
120
121
122
123
124
125
  }
  return(mergedDF)
  
}

#replace blank cells with NAs
removeBlanks <- function(input){
  input <- sapply(input, function(f){is.na(f)<-which(f == '');f})
  return(input)
}