Skimming through some of the high dimension data work, especially in the areas of diseases, environment e.t.c. It seems the high level of computations involved is an eye closer to a lot of questions, especially the possible problem of MULTICOLLINEARITY in the analysis. What is you own take?