Title: Error in compute_pca_impute_missing Function: "could not find function 'estim_ncpPCA'"
Description:
I encountered an error while using the compute_pca_impute_missing
function in R. The error message states, "could not find function 'estim_ncpPCA'".
Code and Function Explanation:
compute_pca_impute_missing(pca_vars, dhs_all_df, save_model = FALSE, model_path = NULL)
Function
function(pca_vars,
dhs_all_df,
save_model = F,
model_path = NULL){
pca_df <- dhs_all_df %>%
dplyr::select(all_of(pca_vars)) %>%
mutate_all(as.numeric) %>%
mutate_all(. %>% scales::rescale(to = c(0,1)))
# file:///Users/robmarty/Downloads/v70i01.pdf
# Estimate number of dimensions; takes a while, so take a random sample
N_use <- ceiling(nrow(pca_df)*0.1)
ncomp <- pca_df %>%
arrange(runif(n())) %>%
head(N_use) %>%
estim_ncpPCA()
#ncomp <- estim_ncpPCA(pca_df)
res.imp <- imputePCA(pca_df, ncp = ncomp$ncp)
res.pca <- prcomp(res.imp$completeObs, scale = T)
out <- res.pca$x[,1]*(-1)
if(save_model){
saveRDS(res.pca, model_path)
}
return(out)
}
This function computes Principal Component Analysis (PCA) and imputes missing values in a dataframe (dhs_all_df
) based on specified variables (pca_vars
). It internally utilizes the estim_ncpPCA
function from the pcaMethods
package for estimating the number of dimensions for PCA.
Error Message:
Error in estim_ncpPCA(.) : could not find function "estim_ncpPCA"
Request for Assistance:
I'm seeking guidance on resolving this error. Specifically, I would like to know how to ensure that the estim_ncpPCA
function is accessible within my R environment.
Thank you for your assistance!