Index
mujpy is here. Documentation on mujpy at ReadTheDocs
Old page: What would be the case for a MuPy (umpy) environment?
- Start from a separate wiki
- Start without gui, edit parameter list, gui later
- The interface should have the following blocks (musrfit-style)
- Data input
- Function assembly
- Parameters
- Plot
- Statistics
- Fourier
- Fit: start from the parameter list of a global fit in a given list of runs
- Fit with different modules: minuit, minuit2, scipy.leastsq, ...
- Standard asymmetry fit
- Tentative Rotating Frame fit (has artifacts)
- Fit globally, across detectors and runs (physical parameters)
- Single detector Minuit fits, along directions α=x,y,z, according to
Ni(t)=N0ie−t/τμ(1±AiPα(t))+Bi
- Polarization projection functions may be reconstructed in two determinations as
Pα,±=±1Ai(Ni−BiN0ie−t/τμ−1)
- Error propagation, with sN=√(N) and sN0,sA,sB from the Minuit fit yield the variance on each determination
s2P=(e−t/τμN0A)2[N(t)+s2B+N(t)−BN0s2N0]+(N−BN0e−t/τμ−1)2s2A
- Rename the two polarization determinations P± and their errors s±; plot polarizations along directions α=x,y,z, according to
P(t)=s2−P+(t)+s2+P−(t)s2++s2−
- with error
sP=sqrt2s+s−√s2++s2−
- The global fit strategy implies that parameters will belong to one of these categories:
- N0i,Bi,Ai, to be fitted globally across physical parameters (i.e. runs)
- optional polar angles θ,ϕ,⋯ to relate different polarization projections, according to a specific model (e.g. internal field in a single crystal, muonium etc.), to be also fitted globally across runs
- Run dependent parameters that remain global across detectors
- Global polarization parameters (e.g. temperature independent muon fractions), across both detectors and runs
Index