jasonmhite / geometrydata Goto Github PK

I made an IPython notebook showing how I extracted this data from the problem object, which you can view here.

• location (meters): (158, 98)
• nominal intensity (counts per second): 3.2140e9

Cross section data can be found in cross_sections.dat. These are macroscopic total cross sections, so they effectively dictate the density as well. They are in units of 1 / meter.

They are in the same order as the geometry dumped from the shapefiles! That is, the first value corresponds to the shapefile that is first in [problem object].domain.geometry and so on.

Locations for the detectors are in detector_locs.dat. Units are again in meters and the order of that list is the ordering of the detectors (detector 1 is the first line, 2 is the second etc). There are 10 in total.

All of the detectors used the same values for dwell time, intrinsic efficiency and face area. These are:

• face area (m^2): .005806
• dwell time (seconds): 5
• efficiency (unitless): .62

Observations are in observations.dat. Each row is an experiment, while the column is the detector number (corresponding to the ordering mentioned in the last section.)

Uncertainties are in uncertainties.dat. These are standard deviations, same format as the observations.

There are 10 replicate sets of observations. You can generate more if you need to by simply calling the loaded problem object without arguments (P()). The uncertainties are the square root of the value.

I have included the actual problem specification file I've been using, in spec.pkl. I can't 100% guarantee that it will work and you might need to regenerate your own using the parameters I provided, but I think it will so it's worth trying.