Hippylib

-------------------
- Switch to GitHub Actions for CI
- Introduce a new function `BiLaplacianComputeCoefficients` to estimate the PDE coefficient based on a
prescribed marginal variance and correlation length
- Allow `BiLaplacianPrior` to take spatially varying coefficients as input
- Add support for non-Gaussian noise models
- Introduce utilities to interpolate cartesian data (expressed as `numpy arrays`) on a `dolfin` mesh.

----------------------------------------
- Support for `FEniCS 2019.1.0`
- Modify `PointwiseObservation` so that ordering of observations targets is respected also in parallel.
Setting the flag `prune_and_sort` to `True` restores previous behavior.
- Remove unused input `tol` from `Model.solveFwd`, `Model.solveAdj`, `Model.solveFwdIncremental`, `Model.solveAdjIncremental`
and from related classes.
- Use `argparse` to set parameters in application drivers from command line
- Use `dl.XDMFFile` to export solutions for visualization in Paraview in all application drivers
- Implement accuracy enhanced SVD algorithm in `randomizedSVD.py`
- Add forward UQ capabilities, using Taylor approximations as control variates
- Introduce hIPPYlib's wrappers to `petcs4py.PETSc.KSP`
- Add reduction operations useful when solving different PDEs concurrently on each process
- Increase coverage of unit testing in CI

----------------------------------------
- Reimplement `BiLaplacianPrior` and `MollifiedBiLaplacianPrior` using a more general framework `SqrtPrecisionPDE_Prior`,
which also supports Gaussian vector fields.
- Update the prior distribution in `model_subsurf.py` and `tutorial/3_SubsurfaceBayesian.ipynb` to use Robin boundary conditions to alleviate boundary artifacts.
- Update the data misfit term in `model_ad_diff.py` and `tutorial/4_AdvectionDiffusionBayesian.ipynb` to use discrete observations in space and time.

----------------------------------------
- Bug fix missing `mpi_comm` in `TimeDependentVector`
- Bug fix in the initialization of the global variable `parRandom`

----------------------------------------
- Add new class `GaussianRealPrior` that implements a finite-dimensional Gaussian prior
- Add a `callback` user-defined function that can be called at the end of each inexact Newton CG iteration
- Add a `__version__` and `version_info` attribute to `hIPPYlib`
- Add `setup.py` to (optionally) install `hIPPYlib` via `pip`
- Add deprecation mechanism
- Deprecate `TimeDependentVector.copy(other)` in favor of `TimeDependentVector.copy()` for consistency with `dolfin.Vector.copy`
- Deprecate `_BilaplacianR.inner(x,y)` for consistency with `dolfin.Matrix`
- CI enhancement via build matrix

----------------------------------------
- Update `README.md` and `paper` according to JOSS reviewers's comments
- Add contributing guidelines
- Fix some typos in notebooks (thanks to **Christian Boehm**)