Error corrections

Currently working on a "proper" way to correct the chi-squared values (and hence the sd's) of intensities from Scala - first pass seems that it is much more effective and much quicker than the current run-scala-lots-remerging solution...