Description Usage Arguments Value Examples
View source: R/JACA_mian_functions.R
Classify observations in the test set based on the supplied matrices of canonical vectors W_list
and the training datasets.
This function first fits a linear discriminant analysis model based on W_list
, trainx
and trainz
, then makes predictions
by W_list
, testx
.
1 | predictJACA(W_list, trainx, trainz, testx, posterior = F)
|
W_list |
A list of view-specific matrices of discriminant vectors. Should be the results of |
trainx |
A list of input data matrices that are used to generate the model: in each sublist, samples are rows and columns are features. |
trainz |
An N by K class indicator matrix that are used to generate the model; rows are samples and columns are class indicator vectors with z_k = 1 if observation belongs to class k. |
testx |
A list of input data matrices Predictions will be made using it. |
posterior |
Logical. If TRUE, the function outputs the posterior probabilities for the classes, otherwise it will outputs the class assignments results. |
prediction |
An n by D matrix with predicted group labels for the test set if posterior = FALSE. Columns are predictions made by each dataset seperately. Otherwise it is a list of posterior probabilities for the test set. |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 | set.seed(1)
# Generate class indicator matrix Z
n = 100
Z=matrix(c(rep(1, n),rep(0, 2 * n)), byrow = FALSE, nrow = n)
for(i in 1:n){
Z[i, ] = sample(Z[i, ])
}
# Generate input data X_list
d = 2
X_list = sapply(1:d, function(i) list(matrix(rnorm(n * 20), n, 20)))
# Train the model
result = jacaTrain(Z, X_list, lambda = rep(0.05, 2), verbose = FALSE, alpha= 0.5, rho = 0.2)
# Make predictions by each dataset seperately
zTest = predictJACA(result, X_list, Z, X_list)
# Make predictions by the concatenated dataset
wJoint = list()
wJoint[[1]] = do.call(rbind, result)
xTrain = list()
xTrain[[1]] = do.call(cbind, X_list)
zTestConcatenated = predictJACA(wJoint, xTrain, Z, xTrain)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.