iFad (version 3.0)

ROC_plot: Calculate the AUC (area under curve) and generate ROC plot

Description

This function evaluates the inference performance of the sparse factor analysis model by using different cut-off values to calculate the sensitivity and specificity of the result of the Gibbs sampling algorithm. In this way, it can give the AUC (area under curve) value and plot the receiver-operating-curve.

Usage

ROC_plot(matrixZ1, matrixZ2, matrixZ_chain, 
plot_name, result_file_name, burn)

Arguments

matrixZ1,matrixZ2

The binary matrix indicating whether each entry of loading matrix W1 and W2 is non-zero. For example, matrixZ1[g,k]=1 indicates that matrixW1[g,k] is non-zero, and vice versa. Dim(matrixZ1)=G1*K. Dim(matrixZ2)=G2*K.

matrixZ_chain

The updated matrixZ in each recorded iteration. A list of length 2, for matrixZ1 and matrixZ2 respectively. matrixZ_chain[[1]] and matrixZ2[[2]] are both matrices with dimension A*B, whereas A is the number of recorded iterations and B the length of matrixZ1/Z2. Each row of the matrix corresponds to the vectorized matrixZ1 or matrixZ2 in each iteration.

plot_name

The name of the ".pdf" file of the plot

result_file_name

The name of the ".RData" file which stores the result of the performance evaluation.

burn

The number of iteration taken as the starting point for the inference of the unknown values in the model. For example,if burn=101, samples from the 101-th iteration to the final iteration of the Gibbs sampling process will be averaged for the inference.

Examples

library(Rlab)
library(MASS)
library(coda)
library(ROCR)


data(matrixZ1)
data(matrixZ2)
data(matrixZ_chain)
ROC_plot(matrixZ1, matrixZ2, matrixZ_chain, plot_name="ROC_plot.pdf", 
result_file_name="ROC_result.RData", burn=1)