I have a date-wise sampling for concentration of various bacteria in a indoor site. Can I use PCA (principal component analysis) or PMF (Positive Matrix Factorization) to club them into some groups according to their sources? Is there any other better methods?