Univariate distribution model

A single distribution fit to a record, optionally censored (via analysis_data()) and optionally nonstationary (via trend()). This is the workhorse model behind univariate_analysis().

Usage

model_univariate(
  family,
  data,
  trends = NULL,
  parameters = NULL,
  parameter_values = NULL,
  use_default_flat_priors = NULL
)

Arguments

family

distribution family name; see distribution_names().

data

a numeric vector of observations, or an analysis_data() frame carrying censored observations.

trends

optional trend() object or list of them, making the model nonstationary.

parameters

optional model_parameter() object or list of them, setting bounds, fixed flags, priors, or starting values.

parameter_values

optional numeric vector of all parameter values, applied last.

use_default_flat_priors

logical; set FALSE alongside a custom prior.

Value

An object of class corehydro_model.

See also

univariate_analysis(), model_validate(), model_simulate().

Examples

peaks <- c(12500, 15300, 8900, 22100, 18700, 14200, 9800, 28500, 17400, 11600)
model_univariate("LogPearsonTypeIII", peaks)
#> <corehydro_model> univariate_distribution: LogPearsonTypeIII
#>   data: 10 exact observations

# Censored: two historical floods known only within a range.
d <- analysis_data(
  exact = peaks,
  interval = data.frame(
    index = c(10, 11), lower = c(30000, 26000),
    value = c(35000, 29000), upper = c(40000, 32000)
  )
)
model_univariate("LogPearsonTypeIII", d)
#> <corehydro_model> univariate_distribution: LogPearsonTypeIII
#>   data: <corehydro_data> 10 exact, 2 interval

# Nonstationary: a linear trend on the location parameter.
model_univariate("Normal", peaks, trends = trend("mean", "Linear"))
#> <corehydro_model> univariate_distribution: Normal
#>   data: 10 exact observations
#>   trends: Linear on parameter 1