Flavio

New observables

- Muon-to-electron conversion implemented by Jonathan Kriewald (thanks!)
- The B_c lifetime `tau_Bc` constraining the decay rates B_c→lν

New measurements

- Muon-to-electron conversion and B_c lifetime (see above)
- Updated Belle measurement of R_D^(*) with semi-leptonic tagging
- Updated Belle measurement of R_K^*
- Updated LHCb measurement of R_K
- New Belle limits on `BR(B0->K*emu)` and `BR(B0->K*mue)`
- New LHCb limits on `BR(Bs->mutau,taumu)` and `BR(B0->mutau,taumu)`

Updated parameters

- Λ_b polarization parameter from BESIII (thanks Danny van Dyk)
- Nucleon matrix elements from FLAG 2019

New features

- New plot function `flavio.plots.diff_plot_exp` for plots of measurements of observables depending on a single parameter (like q²) but where the parameter is not binned but constant
- All the `diff_` and `bin_` plot functions gained a new argument `scale_factor` to rescale the observables by a constant factor

Fixes

- Warning caused by the new `pyyaml` release suppressed (thanks Markus Prim & Kilian Lieret)

This release does not bring new observables or major new features, but instead a couple of small improvements and fixes, the most important ones being:
- Several hadronic parameters from lattice QCD have been updated to the [new FLAG 2019 average](http://arxiv.org/abs/arXiv:1902.08191)
- `flavio.statistics.probability.combine_distributions` now supports multivariate PDFs
- The CKM elements used by the `wilson` package for Wilson coefficient basis translation are now set to the values set in flavio (thanks to jackypheno for reporting a related bug, 78)
- Likelihoods in the `flavio.statistics.likelihood` module now cache their observable predictions for repeated calls (thanks peterstangl!)

New observables

- Leptonic kaon decays `BR(KL->mumu)`, `BR(KS->mumu)` (as well as the electronic modes)
- Lepton flavour violating kaon decays `BR(KL->emu,mue)`, `BR(KS->emu,mue)`
- π→μν total rate, `Gamma(pi+->munu)`
- Some additional Z and W pole observables, e.g. the LFU ratio `Rmue(W->lnu)`

Other improvements

- The B→D and B→D* form factor parameters have been improved: while still using the parametrization from [arXiv:1801.01112](https://arxiv.org/abs/1801.01112), the parameters have been updated from a fit where the LCSR input is taken from [arXiv:1811.00983](https://arxiv.org/abs/1811.00983) rather than [arXiv:0809.0222](https://arxiv.org/abs/0809.0222). This leads in particular to a more reliable prediction for R_D and R_D*.
- Various parametric updates, e.g. quark masses are now taken from [lattice QCD](https://arxiv.org/abs/1802.04248).
- Correlation among B meson bag parameters from FLAG now properly taken into account.
- Two-dimensional likelihoods of the CMS and ATLAS measurements of B_s,d→μμ added (thanks Peter Stangl!)


Breaking change

- The CKM parameter `gamma` has been renamed to `delta` to more accuretely reflect the fact that this is treated internally precisely as the Dirac phase δ. Although this is numerically almost identical with the unitarity triangle angle γ, it is not equivalent.

Fixes

- Factor of 2 in `tau->emue` and `tau->muemu` fixed (thanks to Andrzej Buras!)

... and a couple of smaller fixes.

The first minor release in the v1.x series brings a couple of new observables and some other improvements. Many of these improvements where motivated by [smelli](https://github.com/smelli/smelli) and special thanks go to Jason Aebischer, Jacky Kumar, and Peter Stangl.

New observables

- More leptonic LFV τ decays: τ→3e, τ^-→e^-μ⁺μ^-, τ^-→e^-μ⁺e^-, τ^-→μ^-e⁺μ^-
- τ→πl and τ→Kl (implemented by jackypheno)
- K_S→πlν

Other improvements

- Improved treatment of likelihoods of τ LFV limits
- Implemented the B form factors from a combined fit to an improved LCSR calculation and LQCD results in [arXiv:1811.00983](https://arxiv.org/abs/1811.00983). These form factors have been made the default for B→π and B→K.
- Updated all particle masses and widths to PDG 2018 values
- Other parameter updates: f_K and f_π from FLAG; B bag parameters now use the FLAG average for the SM operator
- Consistent treatment of scale dependence of longitudinal meson decay constants (implemented by peterstangl)

Fixes

- A missing factor of 2 in the dipole contribution to τ→3l was fixed (thanks to Andrzej Buras [yes!] for noticing this)

This minor release was necessary due to changes in the `wilson` and `wcxf` libraries that flavio depends on.

This release increments the version number to 1.0 :tada: This does not mean that flavio is "complete", it is just an expression of the relative maturity of the code, as well as a convenient point in time as a short write-up describing the program will appear on the arXiv preprint server soon. Moreover, this release includes loads of new features, most of which have been developed for the new [smelli](https://github.com/smelli/smelli) package that provides a global likelihood in the space of BSM Wilson coefficients and relies heavily on flavio. Special thanks to my collaborators Jason Aebischer, Jacky Kumar, and Peter Stangl.

New observables: electroweak precision tests!

- *Z* production and decays: electroweak precision observables on the *Z* pole, see the [list](https://flav-io.github.io/docs/obs/zdecays-flavourconservingdecays.html)
- LFV *Z* decays: [list](https://flav-io.github.io/docs/obs/zdecays-fcncdecays.html)
- *W* decays and mass: [list](https://flav-io.github.io/docs/obs/wpmdecays-leptonicdecays.html)

These electroweak precision tests were added by David Straub and Peter Stangl. Other observables:

- τ→φl and τ→ρl added by Peter Stangl (cross-checks by Jacky Kumar)
- F_L in B→D*lν
- Anomalous magnetic dipole moments of the electron, muon, and tau

New likelihood module

The new module `statistics.likelihood` allows to construct likelihoods in terms of parameters and Wilson coefficients given theoretical constraints on parameters and experimental measurements of observables. It was developed for the [smelli](https://github.com/smelli/smelli) package that depends on it and is meant to serve as a basis for `statistics.fits` in the future. Right now, the two modules coexist, but the interface of `fits` might be subject to change in the near future.

Other improvements

- A large number of measurements was added
- Various speed optimizations, some of them drastic

Bug fixes

- Matrix element of the ΔF=2 tensor operator corrected (thanks to Matthew Kirk)
- Bounds on B⁰→Kνν corrected (thanks to James Kahn)


Removed features

As announced in [v.30](https://github.com/flav-io/flavio/releases/tag/v0.30), the `io.flha` module has been removed. If you rely on it (even though it is no longer maintained), you can use it as a standalone script downloading it [here](https://github.com/flav-io/flavio/blob/f0242232f652ae2f7b28806a4c1acc9d0022ecf6/flavio/io/flha.py).