ff: Fuzzy forests algorithm

In fuzzyforest: Fuzzy Forests

Description

Fits the fuzzy forests algorithm. Note that a formula interface for fuzzy forests also exists: ff.formula.

Usage

## Default S3 method:
ff(X, y, Z = NULL, module_membership,
  screen_params = screen_control(min_ntree = 500),
  select_params = select_control(min_ntree = 500), final_ntree = 5000,
  num_processors = 1, nodesize, test_features = NULL, test_y = NULL,
  ...)

ff(X, ...)

Arguments

X	A data.frame. Each column corresponds to a feature vectors.
y	Response vector. For classification, y should be a factor. For regression, y should be numeric.
Z	A data.frame. Additional features that are not to be screened out at the screening step.
module_membership	A character vector giving the module membership of each feature.
screen_params	Parameters for screening step of fuzzy forests. See screen_control for details. screen_params is an object of type screen_control.
select_params	Parameters for selection step of fuzzy forests. See select_control for details. select_params is an object of type select_control.
final_ntree	Number of trees grown in the final random forest. This random forest contains all selected features.
num_processors	Number of processors used to fit random forests.
nodesize	Minimum terminal nodesize. 1 if classification. 5 if regression. If the sample size is very large, the trees will be grown extremely deep. This may lead to issues with memory usage and may lead to significant increases in the time it takes the algorithm to run. In this case, it may be useful to increase nodesize.
test_features	A data.frame containing features from a test set. The data.frame should contain the features in both X and Z.
test_y	The responses for the test set.
...	Additional arguments currently not used.

Value

An object of type fuzzy_forest. This object is a list containing useful output of fuzzy forests. In particular it contains a data.frame with a list of selected the features. It also includes a random forest fit using the selected features.

Note

This work was partially funded by NSF IIS 1251151 and AMFAR 8721SC.

References

Conn, D., Ngun, T., Ramirez C.M., Li, G. (2019). "Fuzzy Forests: Extending Random Forest Feature Selection for Correlated, High-Dimensional Data." Journal of Statistical Software, 91(9). doi: 10.18637/jss.v091.i09

Breiman, L. (2001). "Random Forests." Machine Learning, 45(1), 5-32. doi: 10.1023/A:1010933404324

Zhang, B. and Horvath, S. (2005). "A General Framework for Weighted Gene Co-Expression Network Analysis." Statistical Applications in Genetics and Molecular Biology, 4(1). doi: 10.2202/1544-6115.1128

See Also

ff.formula, print.fuzzy_forest, predict.fuzzy_forest, modplot

Examples

#ff requires that the partition of the covariates be previously determined.
#ff is also handy if the user wants to test out multiple settings of WGCNA
#prior to running fuzzy forests.

library(mvtnorm)
gen_mod <- function(n, p, corr) {
  sigma <- matrix(corr, nrow=p, ncol=p)
  diag(sigma) <- 1
  X <- rmvnorm(n, sigma=sigma)
  return(X)
}

gen_X <- function(n, mod_sizes, corr){
  m <- length(mod_sizes)
  X_list <- vector("list", length = m)
  for(i in 1:m){
    X_list[[i]] <- gen_mod(n, mod_sizes[i], corr[i])
  }
  X <- do.call("cbind", X_list)
  return(X)
}

err_sd <- .5
n <- 500
mod_sizes <- rep(25, 4)
corr <- rep(.8, 4)
X <- gen_X(n, mod_sizes, corr)
beta <- rep(0, 100)
beta[c(1:4, 76:79)] <- 5
y <- X%*%beta + rnorm(n, sd=err_sd)
X <- as.data.frame(X)

Xtest <- gen_X(n, mod_sizes, corr)
ytest <- Xtest%*%beta + rnorm(n, sd=err_sd)
Xtest <- as.data.frame(Xtest)

cdist <- as.dist(1 - cor(X))
hclust_fit <- hclust(cdist, method="ward.D")
groups <- cutree(hclust_fit, k=4)

screen_c <- screen_control(keep_fraction = .25,
                           ntree_factor = 1,
                           min_ntree = 250)
select_c <- select_control(number_selected = 10,
                           ntree_factor = 1,
                           min_ntree = 250)

ff_fit <- ff(X, y, module_membership = groups,
             screen_params = screen_c,
             select_params = select_c,
             final_ntree = 250)
#extract variable importance rankings
vims <- ff_fit$feature_list

#plot results
modplot(ff_fit)

#obtain predicted values for a new test set
preds <- predict(ff_fit, new_data=Xtest)

#estimate test set error
test_err <- sqrt(sum((ytest - preds)^2)/n)

fuzzyforest documentation built on March 25, 2020, 5:09 p.m.