Pyleoclim

- re-introduced support for LiPD objects, including loading, export, and mapping
- rigorous time representation, based on [PaCTS](https://doi.org/10.1029/2019PA003632) proposal
- Gaussian kernel method for coarsening time grids (`gkernel`)
- `filter` method for Series objects
- `common_time` method to put various timeseries on a common timescale, using interpolation, binning or Gaussian kernel
- _MultipleSeries_ methods, including stack plots, correlation, filtering, and PCA
- _EnsembleSeries_ methods, including correlation and plotting
- expanded documentation of data structures (classes)
- various bug fixes

This release fixes some bugs and expands on plotting functionalities for Pyleoclim

- enabled unit tests through TravisCI (to facilitate external contributions)
- developed unit tests for all methods
- expanded documentation, now on [readthedocs](https://pyleoclim-util.readthedocs.io/)
- expanded Jupyter notebook tutorials (under "example_notebooks")
- more user-friendly Getting Started section

* Updates to Series method, including outliers detection
* SSA/MSSA/PCA functions

- complete redesign of the code, now centered on `Series` objects and descendants.
- [user interface](https://github.com/LinkedEarth/Pyleoclim_util/blob/master/example_notebooks/pyleoclim_ui_tutorial.ipynb) allowing succinct, expressive calls
- expanded repertoire of spectral and wavelet methods
- inclusion of singular spectral analysis, permitting missing data
- Sphinx [documentation](http://linkedearth.github.io/Pyleoclim_util/) for all functions
- clean, additive [plot styles](https://github.com/LinkedEarth/Pyleoclim_util/blob/master/example_notebooks/plot_styles.ipynb) inspired by Matplotlib [style sheets](https://matplotlib.org/3.3.1/gallery/style_sheets/style_sheets_reference.html).