I have got a dataset that represents around 30 characteristics from a few hundred samples. Some of these characteristics could be condensed into 2 PCs as shown by a PCA. Now I would like to take these 2 PCs (which retains most of the information of 14 characteristics) and generate a model that uses these 2 PCs alongside the remaining characteristics. Is it possible to take the results of the PCA and merge it with the existing data and can this be used to perform a GLM?
As a side note, there are gaps in the data where values were not recorded. From what I understand the PCA skips these rows so to merge the data it would need to take this into consideration.