ABSTRACT
But now we say, “let’s solve for x2 also.” That’s fine, assuming that m=2. When we solve for x2, we must remember that we have removed one degree of freedom
from the set of measurements of the y’s. This same consideration applies each time we want to retrieve another xi. Therefore, we must be sure that we have enough independent measurements (y’s) to account for each physical variable that affects the measurements. There may be a tendency to ignore this fact and try to write n algorithms to retrieve n variables from m independent measurements, where we treat each algorithm separately, and not realize that the procedure is invalid if m<n.
