Sparse Canonical Correlation Analysis for High-Dimensional Mixed Data
Construct a correlation matrix
Estimate latent correlation matrix
Internal mixedCCA function finding w1 and w2 given R1, R2 and R12
Mixed type simulation data generator for sparse CCA
Kendall's tau correlation
Internal data-driven lambda sequence generating function.
Sparse CCA for data of mixed types with BIC criterion
Internal RidgeCCA function
Internal standard CCA function.
Semi-parametric approach for sparse canonical correlation analysis which can handle mixed data types: continuous, binary and truncated continuous. Bridge functions are provided to connect Kendall's tau to latent correlation under the Gaussian copula model. The methods are described in Yoon, Carroll and Gaynanova (2020) <doi:10.1093/biomet/asaa007> and Yoon, Mueller and Gaynanova (2021) <doi:10.1080/10618600.2021.1882468>.