In markwh/rcmodel: Calibrate and Validate Rating Curves for Riverine Load Estimation

knitr::opts_chunk$set(echo = TRUE)

Improving retrieval vignette.

• simplifying data
• Combine wqpData to rcData

Tracking down a warning when simplifying no3Data to purge redundant units.

• I think it comes from the Mode function.
• text is In c("mg/l", "mg/l as N", "mg/l", "mg/l as N", "mg/l", ... : longer object length is not a multiple of shorter object length

debug(Mode)
no3Simple <- wqp_simplifyConc(no3Data, average = "time")

OK, fixed that. It was improperly specified formula for mutate_ (NSE stuff)

Now I'm testing out the creation of rcData objects via makeRcData(simpleConc, simpleFlow)

rcDataList <- makeRcData(simpleConc = no3Simple, simpleFlow = qSimple)
rawDataList <- makeRcData(simpleConc = no3Simple, simpleFlow = qSimple, type = "raw")
length(rcDataList)
nobs <- sapply(rcDataList, nrow)
nbdl <- sapply(rcDataList, function(df) sum(df$is.bdl))

sum(nobs > 30 & nbdl == 0)
min(which(nobs >= 30))

rcDataList <- rcDataList[nobs >= 30 & nbdl == 0]
rawDataList <- rawDataList[nobs >= 30 & nbdl == 0]


foo <- loadest_cal(rcDataList[[1]])
rcDataList[[1]]
glimpse(rawDataList[[1]])
rawDataList[[1]] %>% 
  plot(conc ~ flow, .)


makeModelData(rawDataList[[1]])

Need to debug makeModelData. For some reason it's giving NaN for c.

debug(makeModelData)
makeModelData(rawDataList[[1]])

Hmm. Problem is zeros in concentration data. Not very likely, especially with nitrate. Should treat these as bdl.

• is there any detection limit info?

rawDataList[[1]]

no3Data %>% 
  filter(MonitoringLocationIdentifier == "MDEDAT04_WQX-SID0112",
         CharacteristicName == "Nitrate",
         ResultSampleFractionText == "Dissolved") %>% 
  glimpse()

None. Could put artificial detection limit at minimum nonzero observation. In any case, zeros should be treated as BDL.

OK, fixed that--need to just omit datasets with zeros for now. But now I'm getting NaN flow values (in the raw data)

glimpse(rawDataList[[1]])

qData %>% 
  filter(MonitoringLocationIdentifier == "USGS-01573750",
         ActivityStartDate == "2002-09-19") %>% 
  glimpse()

Fixed that too.

STILL getting not enough data complaint.

rawDataList[[3]] %>% 
  select(conc, flow) %>% 
  print(n = 50)

Which is USGS-01591510?

at1 <- attr(no3Data, "siteInfo")

at1 %>% 
  filter(MonitoringLocationIdentifier == "USGS-01591510")

I should get rid of zero flow like I did NA flows. For the same reason, really (since NAs are BDLs)

markwh/rcmodel documentation built on May 21, 2019, 12:26 p.m.