Update CHANGES.txt

CHANGES.txt

@@ -1 +1,12 @@
-v0.1.1, April 25, 2020 -- Initial release.
+v0.1.1: April 25, 2020 -- Initial release.
+
+
+v0.1.2: May 1, 2020 -- Current release.
+
+   * Resolved ``FutureWarning`` for deprecated ``pandas.np`` module used to fill missing values (i.e. ``nan`` or ``None``): replaced the use of ``value=pd.np.nan`` w/ ``value=np.nan`` in :obj:`pandas.DataFrame.fillna`.
+
+   * Fixed a bug that used the wrong instance method on a :class:`pandas.DatetimeIndex`. To check if a :class:`pandas.DatetimeIndex` value landed on a Saturday or Sunday, incorrect use of ``pandas.DatetimeIndex.weekday_name`` is replaced w/ :meth:`pandas.DatetimeIndex.day_name()`. The aforementioned is required when comparing datasets constructed at a weekly frequency to those from Ken French's online library (e.g. `Weekly Fama-French 3 Factors <https://mba.tuck.dartmouth.edu/pages/faculty/ken.french/Data_Library/f-f_factors.html>`_).
+
+   * Fixed a bug that **did not** exclude data for an `incomplete` final calendar year in the construction of annual datasets. For example, annual datasets constructed from a given start date to the present should end in 2019, not 2020. 
+
+   * Added options to construct portfolios based on sorts not included in Ken French's online data library. A single anomaly characteristic or multiple anomaly characteristics can now be split into 4 (i.e. quartiles), 6, 8, 20, 25, or 100 (i.e. percentiles) buckets.