R: Run linear model with crossvalidation

extract_lm_cv {waterquality}

R Documentation

Run linear model with crossvalidation

Description

The function runs a linear model on a single water quality parameter and a water quality algorithm and conducts a k-folds cross validation, which returns a data frame containing the following: The r^2, p-value, slope, intercept of the global lm model & average r^2, average RMSE, average MAE from the crossvalidated model

Usage

extract_lm_cv(
  parameter,
  algorithm,
  df,
  train_method = "lm",
  control_method = "repeatedcv",
  folds = 3,
  nrepeats = 5
)

Arguments

`parameter`	water quality parameter
`algorithm`	water quality algorithm
`df`	data frame containing the values for parameter and algorithm arguments
`train_method`	A string specifying which classification or regression model to use (Default = "lm"). See ?caret::train for more details
`control_method`	A string specifying the resampling method (Default = "repeatedcv"). See ?caret::trainControl for more details
`folds`	the number of folds to be used in the cross validation model
`nrepeats`	the number of iterations to be used in the cross validation model

Value

A data frame of the model results

References

Johansen, Richard; et al. (2018). Evaluating the portability of satellite derived chlorophyll-a algorithms for temperate inland lakes using airborne hyperspectral imagery and dense surface observations. Harmful Algae. 76. 10.1016/j.hal.2018.05.001.

R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.