Project

General

Profile

FNEX Projects

Unbinned Likelihood vs Binned ChiSq sensitivities

Description: This project involves using the FNEX 2.0 framework to generate sensitivity plots comparing UBL to binned fit contours. The particulars are left up to the assignee, but would likely involve generating contours using many different pseudo-data samples. Note that pseudo-data is generated in FNEX by selecting appropriate options in the .fcl file, and choosing an RNG seed.

Status: assigned
Assigned to: Sebastian
Progress:

Making ART file-level products

Description: By default, a single ART job cannot load events of different types--'type' here meaning data vs MC, or NearDet vs FarDet. This is due to ART's treatment of all event information as trees, and different types of files having different tree structure. A way around this would be to create file-level (rather than event-level or run-level) FNEX event objects. These could then be loaded into the same job, without employing the current workaround of translating FNEX events into TTrees, concatenating files, and then converting these TTrees back into FNEX events.

The assignee to this project would be looking into the nature of ART file-level objects, in order to develop a working understanding of such objects and to work with Jeremy to implement file-level FNEX objects in FNEX 3.0.

Status: Unassigned
Assigned to: Unassigned
Progress:

Shaping in the nue and numu analyses

Description: This project involves testing the effect of a binned/unbinned fit to nue events as represented in some spectrum, vs the first analysis approach of simply counting events. This project can also be extended to extrapolating in different variables, in both nue and numu spectra, to see how this might affect our final contours.

Status: Unassigned
Assigned to: Unassigned
Progress:

Common selection criteria and selection effects

Description: This project requires playing with the nue and numu selection criteria in order to see where we might introduce common cuts (e.g., a common set of fiducial cuts). We would like to see, first and foremost, how dependent event selection might be on this transition to common cuts. To first order, do we select different events; to second order, how much might we have to adjust PIDs / background estimates? The ultimate goal, is to see whether we might simplify common fits by using common cuts, or if this would simply generate excessive additional uncertainties.

Status: Unassigned
Assigned to: Unassigned
Progress:

GENIE knobs, nue vs numu

Description: This project requires involves exploring the 'knobs' available in GENIE, and identifying how/whether these might differently affect the nue and numu analysis GENIE uncertainties.

Status: Assigned
Assigned to: Jeremy
Progress:

Energy resolution, nue vs numu

Description: How might changing the energy resolution affect the nue energy spectrum / selection, vs the numu energy spectrum / selection. Might be best explored with a 2D fit in energy vs energy resolution.

Status: Unassigned
Assigned to: Unassigned
Progress:

The curious case of overlapping nue/numu selection

Description: As far as we are aware, there are no instances of the nue and numu selection criteria choosing the same event. This project would simply verify that this is the case, then play with loosening cuts to see if there are criteria that might cause us to 'double-count' events in this manner.

NuPrism for NOvA?

Description: Can you reweighs the ND as a function of angle to match the FD predicted energy spectrum? Google NuPrism?

Status: Unassigned
Assigned to: Unassigned
Progress: