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

knitr::opts_chunk$set(echo = TRUE)

WQP should be done updating. Redo retrievals and resave objects.

• nope, looks like it's running late.

Meanwhile, average observations over depth

no3Data %>% 
  filter(!is.na(ResultDepthHeightMeasure.MeasureValue),
         ResultDepthHeightMeasure.MeasureValue != 0) %>% 
  glimpse

So nothing that isn't at zero depth.

qData %>% 
  filter(!is.na(ResultDepthHeightMeasure.MeasureValue),
         ResultDepthHeightMeasure.MeasureValue != 0) %>% 
  glimpse

But should be looking at ActivityDepth

hasdepth <- no3Data %>% 
  filter(!is.na(ActivityDepthHeightMeasure.MeasureValue),
         ActivityDepthHeightMeasure.MeasureValue != 0)

glimpse(hasdepth)

How is this defined? I don't believe they're making observations in a river at 18m depth.

hasdepth %>% 
  group_by(MonitoringLocationIdentifier) %>% 
  summarize(mindepth = min(ActivityDepthHeightMeasure.MeasureValue),
            maxdepth = max(ActivityDepthHeightMeasure.MeasureValue),
            meandepth = mean(ActivityDepthHeightMeasure.MeasureValue),
            n = n()) %>% 
  arrange(desc(maxdepth))

summary(as.factor(hasdepth$ActivityDepthHeightMeasure.MeasureUnitCode))

So I'll need to convert the depth in checkUnits as well.

summary(as.factor(hasdepth$ActivityDepthAltitudeReferencePointText))

That's DONE.

Function to check timezones

summary(as.factor(no3Data$ActivityStartTime.TimeZoneCode))

no3Data %>% 
  group_by(ActivityStartTime.TimeZoneCode) %>% 
  summarize(numna = sum(is.na(ActivityStartTime.Time)),
            numReported = n() - numna)

So a large chunk is reported but has no timezone.

no3Data %>% 
  filter(is.na(ActivityStartTime.TimeZoneCode),
         !is.na(ActivityStartTime.Time)) %>% 
  mutate(starttime = as.factor(ActivityStartTime.Time)) %>% 
  summary()

qData %>% 
  filter(is.na(ActivityStartTime.TimeZoneCode),
         !is.na(ActivityStartTime.Time)) %>% 
  mutate(starttime = as.factor(ActivityStartTime.Time)) %>% 
  summary()

All the data missing timezones are reported as 00:00:00.

• Since this is misleading, remove this information
• How to treat datetime column? options:
  1. remove these rows (make NA)
  2. report start and end time as separate columns
  3. encode the information that time is unknown.

I like option 2 the best. Use ActivityEndDateTime.

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