Skip to contents

Plot the results of automatic (CO)VLMC complexity selection

Source: R/plot_tune.R
plot.tune_vlmc.Rd

This function plots the results of tune_vlmc() or tune_covlmc().

Usage

# S3 method for tune_vlmc
plot(
  x,
  value = c("criterion", "likelihood"),
  cutoff = c("quantile", "native"),
  ...
)

# S3 method for tune_covlmc
plot(
  x,
  value = c("criterion", "likelihood"),
  cutoff = c("quantile", "native"),
  ...
)

Arguments

x

a tune_vlmc object

value

the criterion to plot (default "criterion").

cutoff

the scale used for the cut off criterion (default "quantile")

...

additional parameters passed to base::plot()

Value

the tune_vlmc object invisibly

Details

The standard plot consists in showing the evolution of the criterion used to select the model (AIC() or BIC()) as a function of the cut off criterion expressed in the quantile scale (the quantile is used by default to offer a common default behaviour between vlmc() and covlmc()). Parameters can be used to display instead the loglikelihood() of the model (by setting value="likelihood") and to use the native scale for the cut off when available (by setting cutoff="native").

Customisation

The function sets several default before calling base::plot(), namely:

  • type: "l" by default to use a line representation;

  • xlab: "Cut off (quantile scale)" by default, adapted to the actual scale;

  • ylab: the name of the criterion or "Log likelihood".

These parameters can be overridden by specifying other values when calling the function. All parameters specified in addition to x, value and cutoff are passed to base::plot().

Examples

dts <- sample(as.factor(c("A", "B", "C")), 100, replace = TRUE)
tune_result <- tune_vlmc(dts)
## default plot
plot(tune_result)

## likelihood
plot(tune_result, value = "likelihood")

## parameters overriding
plot(tune_result,
  value = "likelihood",
  xlab = "Cut off", type = "b"
)

pc <- powerconsumption[powerconsumption$week %in% 10:12, ]
dts <- cut(pc$active_power, breaks = c(0, quantile(pc$active_power, probs = c(0.5, 1))))
dts_cov <- data.frame(day_night = (pc$hour >= 7 & pc$hour <= 17))
dts_best_model_tune <- tune_covlmc(dts, dts_cov, criterion = "AIC")
plot(dts_best_model_tune)

plot(dts_best_model_tune, value = "likelihood")