Fit a cv_model

cv_model() fits a model.

cv_model(x, ...)

# S3 method for default
cv_model(x, ...)

# S3 method for data.frame
cv_model(x, y, intercept = TRUE, ...)

# S3 method for matrix
cv_model(x, y, intercept = TRUE, ...)

# S3 method for formula
cv_model(formula, data, intercept = TRUE, ...)

# S3 method for recipe
cv_model(x, data, intercept = TRUE, ...)

Arguments

x	Depending on the context: A data frame of predictors. A matrix of predictors. A recipe specifying a set of preprocessing steps created from recipes::recipe().
...	See Details
y	When x is a data frame or matrix, y is the outcome specified as: A data frame with 1 numeric column. A matrix with 1 numeric column. A numeric vector.
intercept	Should intercept column be created? Default is TRUE.
formula	A formula specifying the outcome terms on the left-hand side, and the predictor terms on the right-hand side.
data	When a recipe or formula is used, data is specified as: A data frame containing both the predictors and the outcome.

Value

A bmbstats_cv_model object.

Details

Extra parameters using ... are forwarded to implementation function. These parameters are the following:

model_func

Model function. Default is lm_model. See also baseline_model

predict_func

Predict function. Default is generic_predict

perf_func

Model performance function. Default is performance_metrics

SESOI_lower

Function or numeric scalar. Default is SESOI_lower_func

SESOI_upper

Function or numeric scalar. Default is SESOI_upper_func

control

Control structure using model_control. The parameters used in cv_model are cv_folds, and cv_strata

na.rm

Should NAs be removed? Default is FALSE. This is forwarded to model_func, predict_func, perfr_func, SESOI_lower, and SESOI_upper

In summary, cv_model represents a wrapper function, that performs model_func within the cross-validation loop and provide it's predictive performance metrics using perf_func

Examples

data("vertical_jump_data")

m1 <- cv_model(
  `Post-test` ~ `Pre-test` * Group * `Squat 1RM`,
  vertical_jump_data,
  control = model_control(
    cv_repeats = 10,
    cv_folds = 3,
    cv_strata = vertical_jump_data$Group
  )
)
#> Cross-validating: 3 folds, 10 repeats
#> Done!

m1
#> Training data consists of 8 predictors and 30 observations. Cross-Validation of the model was performed using 10 repeats of 3 folds.
#> 
#> Model performance:
#> 
#>         metric      training training.pooled testing.pooled       mean
#>            MBE -1.563192e-14    1.894792e-15      0.1826778  0.1826778
#>            MAE  8.763836e-01    7.121335e-01      1.4909445  1.4909445
#>           RMSE  1.059620e+00    9.051436e-01      2.0197935  1.9203781
#>           PPER  7.787868e-01    8.561700e-01      0.4872447  0.5048117
#>  SESOI to RMSE  2.543190e+00    2.929626e+00      1.3128731  1.5571820
#>      R-squared  9.744091e-01    9.813267e-01      0.9077786  0.8959661
#>         MinErr -2.117309e+00   -2.405189e+00     -4.3581655 -2.7076806
#>         MaxErr  1.706648e+00    2.053378e+00     10.0065030  3.5323611
#>      MaxAbsErr  2.117309e+00    2.405189e+00     10.0065030  4.1730050
#>         SD        min        max
#>  0.7000498 -0.8258230  1.7834543
#>  0.4482005  0.5886351  2.4593054
#>  0.6365696  0.7358860  3.6568184
#>  0.1343872  0.2765989  0.8566607
#>  0.5712633  0.7718072  3.3785747
#>  0.1125430  0.4421786  0.9903590
#>  1.0021786 -4.3581655 -0.7893271
#>  2.3255646  0.4105829 10.0065030
#>  1.9372322  1.2929473 10.0065030
plot(m1, "residuals")