This is a collection of simple python scripts to calculate the sensitivity for a basic counting experiment under different interpretations of radioassay results.

Author: Raymond Tsang

1. Python 3.4 (recommended) or 2.7
2. PyROOT (for histogramming and Feldman-Cousins calculator)
3. numpy (for random number generation and miscellaneous numerical utilities)

Examples of user code can be found in main.py.

First, for each part of the Detector in the actual experiment, the user needs to specify, the true specific activity, the hit efficiency and the mass. For example, for ncomp identical parts:

from detector import Detector, Component
comps = []
spec_act = 1e-6 # Specific activity of the part in Bq/kg
eff = 0.1       # Hit efficiency of the part in the actual detector
mass = 1.       # Mass of the part in kg
for ic in range(ncomp):
  comps.append(Component(spec_act,eff,mass))
det = Detector(comps)

Then define the detector for radioassay, e.g. an HPGe detector.:

from facilities import HPGe
ge = HPGe(10./86400.)  # Background rate of HPGe in counts per second

Let the detector perform an assay measurement and associate it with the assayed component:

from assay import Assay
assayed_mass = 1.   # Mass of part being assayed in kg
result = ge.count(comp.trueimp,assayed_mass,livetime=14*86400)
comp.assay = Assay(result)

Here Assay performs random draws from the selected Bayesian prior.

Finally perform a sensitivity calculation:

livetime = 10*365*86400 # Experiment livetime in seconds
method = 'Uniform'      # Way to interpret limits
nsenstoys = 10000       # Number of toys used in sensitivity calculation
usetruth = False        # If True, use true radioassay values. If False, use assayed values and interpret according to "method".
sens, uls = calc_sens(det, method, livetime, nsenstoys, usetruth)

It will return the 90% C.L. upper limits (uls) and their mean (sens)

Available priors (method):

(See assay.py)

python main.py <ntoys> <true_lambda> <ncomp> <spec_act> <livetime> <method> <nsenstoys>

1. ntoys: Number of toy MCs for the assay procedure
2. true_lambda: Expected true total background rate
3. ncomp: Number of parts
4. spec_act: Specific activity of every part
5. livetime: Livetime of the experiment
6. method: Method to interpret assay results
7. nsenstoys: Number of toy MCs used in each sensitivity calculation