Skip to contents

Fit machine learning with the consolidated scoped result contract

Source: R/stats_compat.R
fit_ml_panel.Rd

fit_ml_panel() is the unambiguous positional entry point for code migrated from idiostats. It always returns the consolidated scoped result used by metrics(), predictions(), tuning(), and the result-view functions. Named fit_ml(data, y = ..., x = ..., id = ...) calls are equivalent.

Usage

fit_ml_panel(data, y, x, id, ...)

Arguments

data

A data frame or matrix.

y

Outcome column name.

x

Predictor selector.

id

Person identifier column.

...

Consolidated ML controls; see fit_ml().

Value

An idiographic_fit retaining the idiostats_fit compatibility class.

Examples

fit <- fit_ml_panel(srl, "effort", "efficacy:monitoring", "name",
                    time = "day", scope = "pooled", model = "ridge")
metrics(fit, overall = TRUE)
#>    scope model estimator  subject subgroup    n     rmse      mae      bias
#> 1 pooled ridge    native .overall     .all 1150 20.18719 16.28609 0.7323493
#>   r_squared
#> 1  0.435318