Merge pull request #138 from njsmith/release-0.5.1

Release engineering for v0.5.1

Author: njsmith

doc/changes.rst

@@ -3,6 +3,15 @@ Changes
 
 .. currentmodule:: patsy
 
+v0.5.1
+------
+
+* The Python 3.6.7 and 3.7.1 point releases changed the standard
+  tokenizer module in a way that broke patsy. Updated patsy to work
+  with these point releases. (See `#131
+  <https://github.com/pydata/patsy/pull/131>`__ for details.)
+
+
 v0.5.0
 ------
 

doc/library-developers.rst

@@ -120,6 +120,7 @@ And here's how it can be used:
        exec(f.read())
 
 .. ipython:: python
+   :okwarning:
 
    from patsy import demo_data
    data = demo_data("x", "y", "a")

doc/quickstart.rst

@@ -43,6 +43,7 @@ method attached, so we can pass them directly to a regression function
 like :func:`np.linalg.lstsq`:
 
 .. ipython:: python
+   :okwarning:
 
    outcome, predictors = dmatrices("y ~ x1 + x2", data)
    betas = np.linalg.lstsq(predictors, outcome)[0].ravel()

doc/stateful-transforms.rst

@@ -151,7 +151,7 @@ careful of. Let's put them in red:
 .. warning::
 
    If you are unwise enough to ignore this section, write a function
-   like `naive_center` above, and use it in a formula, then Patsy will
+   like ``naive_center`` above, and use it in a formula, then Patsy will
    not notice. If you use that formula with :func:`incr_dbuilders` or
    for predictions, then you will just silently get the wrong
    results. We have a plan to detect such cases, but it isn't

patsy/version.py

@@ -17,4 +17,4 @@
 # want. (Contrast with the special suffix 1.0.0.dev, which sorts *before*
 # 1.0.0.)
 
-__version__ = "0.5.0+dev"
+__version__ = "0.5.1+dev"