predict_DO: Predict DO from a fitted model.

In USGS-R/streamMetabolizer: Models for Estimating Aquatic Photosynthesis and Respiration

Predict DO from a fitted model.

Description

A function in the metab_model_interface. Returns predictions of dissolved oxygen.

Usage

predict_DO(
  metab_model,
  date_start = NA,
  date_end = NA,
  ...,
  attach.units = deprecated(),
  use_saved = TRUE
)

## S3 method for class 'metab_Kmodel'
predict_DO(metab_model, date_start = NA, date_end = NA, ..., use_saved = TRUE)

## S3 method for class 'metab_model'
predict_DO(
  metab_model,
  date_start = NA,
  date_end = NA,
  ...,
  attach.units = deprecated(),
  use_saved = TRUE
)

## S3 method for class 'metab_night'
predict_DO(metab_model, date_start = NA, date_end = NA, ..., use_saved = TRUE)

## S3 method for class 'metab_sim'
predict_DO(metab_model, date_start = NA, date_end = NA, ...)

Arguments

metab_model	A metabolism model, implementing the metab_model_interface, to use in predicting metabolism
date_start	Date or a class convertible with as.Date. The first date (inclusive) for which to report DO predictions. If NA, no filtering is done.
date_end	Date or a class convertible with as.Date. The last date (inclusive) for which to report DO predictions. If NA, no filtering is done.
...	Other arguments passed to class-specific implementations of predict_DO
attach.units	(deprecated, effectively FALSE in future) logical. Should units be attached to the output?
use_saved	logical. Is it OK to use predictions that were saved with the model?

Value

A data.frame of dissolved oxygen predictions at the temporal resolution of the input data

Methods (by class)

• metab_Kmodel: Throws an error because models of type 'Kmodel' can't predict DO. metab_Kmodel predicts K at daily timesteps and usually knows nothing about GPP or ER. So it's not possible to predict DO from this model. Try passing the output to metab_mle and THEN predicting DO.

• metab_model: This implementation is shared by many model types

• metab_night: Generate nighttime dissolved oxygen predictions from a nighttime regression model. metab_night only fits ER and K, and only for the darkness hours, so predictions are only generated for those hours.

• metab_sim: Simulate values for DO.obs (with process and observation error), DO.mod (with process error only), and DO.pure (with no error). The errors are randomly generated on every new call to predict_DO.

See Also

Other metab_model_interface: get_data_daily(), get_data(), get_fitting_time(), get_fit(), get_info(), get_param_names(), get_params(), get_specs(), get_version(), predict_metab()

Examples

dat <- data_metab('3', day_start=12, day_end=36)
mm <- metab_night(specs(mm_name('night')), data=dat)
preds <- predict_DO(mm, date_start=get_fit(mm)$date[3])
head(preds)

USGS-R/streamMetabolizer documentation built on Aug. 15, 2023, 7:50 a.m.