Based on review

Author: springcoil

docs/source/api.rst

@@ -16,3 +16,4 @@ API Reference
    api/diagnostics
    api/backends
    api/math
+   api/data

docs/source/api/inference.rst

@@ -65,3 +65,11 @@ ADVI minibatch
 
 .. automodule:: pymc3.variational.advi_minibatch
    :members:
+
+ADVI approximations
+^^^^^^^^^^^^^^^^^^^
+
+.. currentmodule:: pymc3.variational.approximations
+
+.. automodule:: pymc3.variational.approximations
+   :members:

docs/source/api/sampling.rst

@@ -150,29 +150,25 @@ def sample(draws, step=None, init='ADVI', n_init=200000, start=None,
     trace : pymc3.backends.base.MultiTrace
         A `MultiTrace` object that contains the samples.
 
-        Examples
+    Examples
     --------
     .. code:: ipython
 
         >>> import pymc3 as pm
-        ... import numpy as np
-        ... np.random.seed(20090425)
-        ... n = 20
-        ... X = np.sort(3*np.random.rand(n))
-
-    Now let's set up a model and sample from it.
+        ... n = 100
+        ... h = 61
+        ... alpha = 2
+        ... beta = 2
 
     .. code:: ipython
-        >>> with pm.Model() as model:
-        ...  b0 = pm.Normal('b0', mu=0, sd=100)
-        ...  yest = b0 * X
-        ...
-        >>> with model:
-        ...      trc = pm.sample(100)
-        ...
-        >>> pm.df_summary(trc)
-                mean         sd  mc_error     hpd_2.5    hpd_97.5
-        b0  7.004575  76.547562  7.654756 -161.098322  136.327346
+
+        >>> with pm.Model() as model: # context management
+        ...     p = pm.Beta('p', alpha=alpha, beta=beta)
+        ...     y = pm.Binomial('y', n=n, p=p, observed=h)
+        ...     trace = pm.sample(2000, tune=1000, njobs=4)
+        >>> pm.df_summary(trace)
+               mean        sd  mc_error   hpd_2.5  hpd_97.5
+        p  0.604625  0.047086   0.00078  0.510498  0.694774
     """
     model = modelcontext(model)