Implementation of flux fit in Cafana - To a large extent, it is - - PowerPoint PPT Presentation

Implementation of flux fit in Cafana

• To a large extent, it is done.

Cafana oscillation

• It builds a two dimensional histogram when

loading trees, true energy vs. desired variable.

• Each time when oscillates, the weights in above

two dimensional histogram changes so that the desired variable spectrum changes based on true neutrino energy.

Flux fit

• Register a loader to a prediction.
• Added a off-axis prediction class to apply cuts to

the spectrum in order to select different off-axis Samples.

• Run the loader with cuts above to have each
• ff-axis prediction filled.
• When fit, Cafan provides oscillation curve,

input to and run Luke’s flux tool to obtain c values and apply to each of the off-axis sample.

Few comments

• The flux fit is in the format of -0.25m-1.75m,

1.75m-3.75m and so on with 2 m each sample. It seems not very flexible. Luke will change in the future!?

• This makes a little tricky to adjust my CAF samples.

But it should be OK.

• Sorry! I did not generate sample at 0 ! To be

consistent, I am doing it. I can only use 4 m off-axis As at center.

• Luke and I may use different flux, this is a problem.

I may try to put my flux in Luke’s tool, is it easy, Luke?

Some validations

C values no systematics numu FHC Off-axis sample each in 2 m span Off-axis sample each in 2 m span Cafana best fit c values Pure Luke’s tool

Some validations

Cafana oscillation spectrum (our target spectrum) With flux fit – not quite same

• Ignore the rate, please.
• We are using different fluxes. Will try to

Put consistent flux in CAF sample and Flux fit tool!