Elong-13-05-31-b

Problems in Rates Code

In redoing the x-rebinning part of the rates code so that the number of counts that come from each setting for a particular x bin were accounted for, I found a few problems in the rates code which have been fixed.

Problem 1: I was over-counting $\mathcal{L}\sigma$. This was counted for each incremental theta bin, then multiplied by the size of the increment, then the rest of the acceptance and the time (in hours). The total should have been zero'ed-out after each loop around dE, but that didn't happen. This meant that it just kept adding more and more of them instead of just for what was accurate for the dE bin. This was corrected.

Problem 2: The incremental theta acceptance was calculated as $2d\theta$ where $\theta$ was in degrees. Everything else was in radians, so to fix the mis-match, it was multiplied by $\frac{\pi}{180}$. In addition, the increment includes the entire $\theta$ range, so the 2 in front was also removed.

Problem 3: The rate calculation was in Hz, but was being multiplied by hours. A factor of 3600 was included to fix this.

With these three problems fixed, the uncertainties in the rebinned x ranges are:

Old, BAD Uncertainties

For reference only, the plot below shows what the calculations were getting before the corrections were applied.

--E. Long 19:49, 3 June 2013 (UTC)