Introduction to manager

Justin Jent

2020-03-08

Source: vignettes/intro.Rmd

The manager packages provides functions for reading external data.

data <- read_manages3("C:/path/to/manages/Site.mdb")

Example data sets for groundwater in Indiana and Ohio are provided with the package

data("gw_data")
data("ohio_data")
data("indiana_data")

Prediction Intervals

data("gw_data")

wells <- c("MW-1", "MW-2", "MW-3", "MW-4")

params <- c("Sulfate, total",
            "Arsenic, dissolved",
            "Boron, dissolved")

background <- lubridate::ymd(c("2007-12-20", "2012-01-01"), tz = "UTC")

# first group data by location, param, and background
# estimate percent less than
background_data <- gw_data %>%
  filter(location_id %in% wells, param_name %in% params,
         sample_date >= background[1] & sample_date <= background[2]) %>%
  group_by(location_id, param_name, default_unit) %>%
  percent_lt() %>%
  est_dist(., keep_data_object = TRUE) %>%
  arrange(location_id, param_name)

pred_int <- background_data %>%
  mutate(pred_int = case_when(
    distribution == "Normal" ~ map(.x = data,
                                   ~EnvStats::predIntNorm(x = .x$analysis_result)),
    distribution == "Lognormal" ~ map(.x = data,
                                      ~EnvStats::predIntLnorm(x = .x$analysis_result)),
    distribution == "Nonparametric" ~ map(.x = data,
                                          ~EnvStats::predIntNpar(x = .x$analysis_result))
    )
  )

pred_int_table <- pred_int %>%
  mutate(distribution = distribution,
         sample_size = map(.x = pred_int, ~ .x$sample.size),
         lpl = map(.x = pred_int, ~ .x$interval$limits["LPL"]),
         upl = map(.x = pred_int, ~ .x$interval$limits["UPL"]),
         conf_level = map(.x = pred_int, ~ .x$interval$conf.level)) %>%
  select(-data, -pred_int) %>%
  unnest()
## Warning: `cols` is now required.
## Please use `cols = c(sample_size, lpl, upl, conf_level)`
kable(pred_int_table)
location_id default_unit param_name distribution sample_size lpl upl conf_level
MW-1 ug/L Arsenic, dissolved Nonparametric 14 1.1000000 10.0000000 0.8666667
MW-1 mg/L Boron, dissolved Nonparametric 14 0.0800000 1.3700000 0.8666667
MW-1 mg/L Sulfate, total Nonparametric 14 15.0000000 117.0000000 0.8666667
MW-2 ug/L Arsenic, dissolved Nonparametric 14 1.0000000 10.0000000 0.8666667
MW-2 mg/L Boron, dissolved Lognormal 14 1.7298714 2.5721482 0.9500000
MW-2 mg/L Sulfate, total Nonparametric 14 158.0000000 200.0000000 0.8666667
MW-3 ug/L Arsenic, dissolved Nonparametric 14 0.4000000 10.0000000 0.8666667
MW-3 mg/L Boron, dissolved Lognormal 14 1.0454563 1.8718412 0.9500000
MW-3 mg/L Sulfate, total Lognormal 14 109.0915641 163.9923875 0.9500000
MW-4 ug/L Arsenic, dissolved Normal 14 -0.0334501 12.8334501 0.9500000
MW-4 mg/L Boron, dissolved Normal 14 -0.0616289 0.8947718 0.9500000
MW-4 mg/L Sulfate, total Nonparametric 14 22.1000000 120.0000000 0.8666667

Confidence Intervals

conf_int <- background_data %>%
  mutate(conf_int = case_when(
    distribution == "Normal" ~ map(.x=data,
                                   ~EnvStats::enorm(x = .x$analysis_result,
                                          ci = TRUE, ci.type = "lower",
                                          conf.level = 0.99,
                                          ci.param = "mean")),
    distribution == "Lognormal" ~ map(.x = data,
                                      ~EnvStats::elnormAlt(x = .x$analysis_result,
                                                 ci = TRUE, ci.type = "lower",
                                                 ci.method = "land",
                                                 conf.level = 0.99)),
    distribution == "Nonparametric" ~ map(.x = data,
                                          ~EnvStats::eqnpar(x = .x$analysis_result,
                                                  ci = TRUE, ci.type = "lower",
                                                  ci.method = "interpolate",
                                                  approx.conf.level = 0.99))
    )
  )

conf_int_table <- conf_int %>%
  mutate(distribution = distribution,
         sample_size = map(.x = conf_int, ~ .x$sample.size),
         lcl = map(.x = conf_int, ~ round(.x$interval$limits["LCL"], 3)),
         ucl = map(.x = conf_int, ~ .x$interval$limits["UCL"]),
         conf_level = map(.x = conf_int, ~ .x$interval$conf.level)) %>%
  select(-data, -conf_int) %>%
  unnest()
## Warning: `cols` is now required.
## Please use `cols = c(sample_size, lcl, ucl, conf_level)`
kable(conf_int_table)
location_id default_unit param_name distribution sample_size lcl ucl conf_level
MW-1 ug/L Arsenic, dissolved Nonparametric 14 1.500 Inf 0.99
MW-1 mg/L Boron, dissolved Nonparametric 14 0.265 Inf 0.99
MW-1 mg/L Sulfate, total Nonparametric 14 50.409 Inf 0.99
MW-2 ug/L Arsenic, dissolved Nonparametric 14 1.382 Inf 0.99
MW-2 mg/L Boron, dissolved Lognormal 14 1.993 Inf 0.99
MW-2 mg/L Sulfate, total Nonparametric 14 162.818 Inf 0.99
MW-3 ug/L Arsenic, dissolved Nonparametric 14 0.600 Inf 0.99
MW-3 mg/L Boron, dissolved Lognormal 14 1.292 Inf 0.99
MW-3 mg/L Sulfate, total Lognormal 14 126.169 Inf 0.99
MW-4 ug/L Arsenic, dissolved Normal 14 4.362 Inf 0.99
MW-4 mg/L Boron, dissolved Normal 14 0.265 Inf 0.99
MW-4 mg/L Sulfate, total Nonparametric 14 78.510 Inf 0.99

The manager pacakges simplifies the above steps.

manager_pred_int <- background_data %>% pred_int(.)
## Warning: `cols` is now required.
## Please use `cols = c(sample_size, lpl, upl, conf_level)`
kable(manager_pred_int)
location_id default_unit param_name distribution sample_size lpl upl conf_level
MW-1 ug/L Arsenic, dissolved Nonparametric 14 -Inf 10.0000000 0.9333333
MW-1 mg/L Boron, dissolved Nonparametric 14 -Inf 1.3700000 0.9333333
MW-1 mg/L Sulfate, total Nonparametric 14 -Inf 117.0000000 0.9333333
MW-2 ug/L Arsenic, dissolved Nonparametric 14 -Inf 10.0000000 0.9333333
MW-2 mg/L Boron, dissolved Lognormal 14 0 2.4818064 0.9500000
MW-2 mg/L Sulfate, total Nonparametric 14 -Inf 200.0000000 0.9333333
MW-3 ug/L Arsenic, dissolved Nonparametric 14 -Inf 10.0000000 0.9333333
MW-3 mg/L Boron, dissolved Lognormal 14 0 1.7761065 0.9500000
MW-3 mg/L Sulfate, total Lognormal 14 0 158.0765513 0.9500000
MW-4 ug/L Arsenic, dissolved Normal 14 -Inf 11.6737350 0.9500000
MW-4 mg/L Boron, dissolved Normal 14 -Inf 0.8085698 0.9500000
MW-4 mg/L Sulfate, total Nonparametric 14 -Inf 120.0000000 0.9333333 
manager_conf_int <- background_data %>%
  conf_int(., ci_type = "lower", conf_level = 0.99)
## Warning: `cols` is now required.
## Please use `cols = c(sample_size, lcl, ucl, conf_level)`
kable(manager_conf_int)
location_id default_unit param_name distribution sample_size lcl ucl conf_level
MW-1 ug/L Arsenic, dissolved Nonparametric 14 1.500 Inf 0.99
MW-1 mg/L Boron, dissolved Nonparametric 14 0.265 Inf 0.99
MW-1 mg/L Sulfate, total Nonparametric 14 50.409 Inf 0.99
MW-2 ug/L Arsenic, dissolved Nonparametric 14 1.382 Inf 0.99
MW-2 mg/L Boron, dissolved Lognormal 14 1.993 Inf 0.99
MW-2 mg/L Sulfate, total Nonparametric 14 162.818 Inf 0.99
MW-3 ug/L Arsenic, dissolved Nonparametric 14 0.600 Inf 0.99
MW-3 mg/L Boron, dissolved Lognormal 14 1.292 Inf 0.99
MW-3 mg/L Sulfate, total Lognormal 14 126.169 Inf 0.99
MW-4 ug/L Arsenic, dissolved Normal 14 4.362 Inf 0.99
MW-4 mg/L Boron, dissolved Normal 14 0.265 Inf 0.99
MW-4 mg/L Sulfate, total Nonparametric 14 78.510 Inf 0.99