myaccuracy: Functions to assess model performance

In lampk/R306:

Description

Functions to assess model performance

Usage

myaccuracy(predicted, y, criteria = mse, ...)

mse(predicted, y)

loglik_binary(predicted, y)

Brier(predicted, y, confint = FALSE)

dis_slope(predicted, y)

c_index(predicted, y, confint = FALSE, ci_collapse = "-",
  ci_parentheses = FALSE)

concordance(x, y)

calib(predicted, y, confint = FALSE)

binperform(predicted, y, cutoff = 0.5)

sen(predicted, y, cutoff)

spe(predicted, y, cutoff)

ppv(predicted, y, cutoff)

npv(predicted, y, cutoff)

acc(predicted, y, cutoff)

auc(predicted, y, alpha = 0.05, digits = 2)

best.youden(predicted, y, by = 0.01, digits = 2, name = "")

best.acc(predicted, y, by = 0.01, digits = 2, name = "")

new.accuracy(predicted, y, cutoff = 0.5)

Arguments

predicted	Predicted values from fitted models
y	Response

Value

numeric value of model's performance

Functions

• mse: Mean squared errors

• loglik_binary: log-likelihood (binary case)

• Brier: Brier score

• dis_slope: Discrimination slope

• c_index: AUC (with confidence interval)

• concordance: Concordance

• calib: Calibration slope and intercept (with confidence intervals)

• binperform: Other measures of binary test

• sen: Sensitivity

• spe: Specificity

• ppv: Positive predictive value

• npv: Negative predictive value

• acc: Accuracy rate

• auc: AUC

• best.youden: Best Youden's index

• best.acc: Best accuracy rate

• new.accuracy: Accuracy of binary test

lampk/R306 documentation built on May 20, 2019, 7:34 p.m.