 |
 |
Thank you very much for your answer! I would have not arrived to this conclusion on my own. I believe I can implement the equation provided. Just as a point of clarification, when you said: "You would need to feed the pseudo-inverse of that leading matrix into fsl_glm to denoise the data", the equation would be: fsl_glm_matrix = pinv(( Id - X_2 * (0 | Id_2) pinv(X)) * Y) at least in its non-compact form? "where n2 is the number of nice components" : was nice supposed to be noise components? again thank you for the response, James