confusion_matrix: Calculate various statistics from a confusion matrix

In signaturescience/skater: Utilities for SNP-Based Kinship Analysis

Calculate various statistics from a confusion matrix

Description

Given a vector of predictions and target values, calculate numerous statistics of interest. Modified from m-clark/confusion_matrix.

Usage

confusion_matrix(
  prediction,
  target,
  positive = NULL,
  prevalence = NULL,
  dnn = c("Predicted", "Target"),
  longer = FALSE,
  ...
)

Arguments

prediction	A vector of predictions
target	A vector of target values
positive	The positive class for a 2-class setting. Default is NULL, which will result in using the first level of target.
prevalence	Prevalence rate. Default is NULL.
dnn	The row and column headers for the contingency table returned. Default is 'Predicted' for rows and 'Target' for columns.
longer	Transpose the output to long form. Default is FALSE (requires tidyr 1.0).
...	Other parameters, not currently used.

Details

This returns accuracy, agreement, and other statistics. See the functions below to find out more. Originally inspired by the confusionMatrix function from the caret package.

Value

A list of tibble(s) with the associated statistics and possibly the frequency table as list column of the first element. If classes contain >1 numeric class and a single non-numeric class (e.g., "1", "2", "3", and "Unrelated", the RMSE of the reciprocal of the Targets + 0.5 will also be returned.)

References

Kuhn, M., & Johnson, K. (2013). Applied predictive modeling.

See Also

calc_accuracy calc_stats

Examples

prediction = c(0,1,1,0,0,1,0,1,1,1)
target     = c(0,1,1,1,0,1,0,1,0,1)
confusion_matrix(prediction, target, positive = '1')

set.seed(42)
prediction = sample(letters[1:4], 250, replace = TRUE, prob = 1:4)
target     = sample(letters[1:4], 250, replace = TRUE, prob = 1:4)
confusion_matrix(prediction, target)

prediction = c(rep(1, 50), rep(2, 40), rep(3, 60))
target     = c(rep(1, 50), rep(2, 50), rep(3, 50))
confusion_matrix(prediction, target)
confusion_matrix(prediction, target) %>% purrr::pluck("Table")
confusion_matrix(prediction, target, longer=TRUE)
confusion_matrix(prediction, target, longer=TRUE) %>%
  purrr::pluck("Other") %>%
  tidyr::spread(Class, Value)

# Prediction with an unrelated class
prediction = c(rep(1, 50), rep(2, 40), rep(3, 60), rep("Unrelated", 55))
target     = c(rep(1, 50), rep(2, 50), rep(3, 55), rep("Unrelated", 50))
confusion_matrix(prediction, target)
# Prediction with two unrelated classes
prediction = c(rep(1, 50), rep(2, 40), rep("Third", 60), rep("Unrelated", 55))
target     = c(rep(1, 50), rep(2, 50), rep("Third", 55), rep("Unrelated", 50))
confusion_matrix(prediction, target)

signaturescience/skater documentation built on Feb. 11, 2023, 4:19 p.m.