View source: R/deconvolute_ccl.R
Finds the MSE for a NMF fit given a rank of 0 to 2. If the MSE is lowest for rank 0, then this suggests that the design matrix is complete and perfectly explains the observed data. If the MSE for a rank 1 or rank 2 is lowest, this suggests that the design matrix partially explains the observed data.
Upon examining the H matrix, we can see the VERY rough proportion of the data in rank k that is unexplained from the design matrix.
1 | checkMse(M, M0)
|
M |
Input matrix |
M0 |
Design matrix |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.