I would like to perform matrix factorization in R on my data. However, it is a special case and I couldn't find any package to perform what I am looking for. I would be thankful if anybody knows such package/method in R to introduce. So, my issue is as follows: Lets say y is the matrix that I have. So, I am looking for y1 and y2 as y1*y2 generates the y as any matrix factorization works. The only difference is that I want to fill some of the columns of y1 with fixed numbers and only do factorization for the last column of y1 and the whole y2. (as below figure, I have fixed columns with predetermined values (just splited y1 to show some columns are determined except last column), and looking for finding the predicted ones)
I hope it is clear, enough. I would appreciate any help in this.
Some R packages, like Singular Value Decomposition (SVD) and Non-Negative Matrix Factorization (NMF), allow you to perform "Matrix Factorization." You can handle your special case by altering the data matrix "y" by filling in the columns with predetermined values and then using the package's NMF or SVD function to factorize the modified data matrix. The final column of y1 and the entirety of y2 can be extracted following factorization.