Build docs on RTD using conda

doc/environment.yml

@@ -0,0 +1,10 @@
+name: xray
+dependencies:
+  - python=2.7
+  - numpy=1.10
+  - pandas=0.17.1
+  - numpydoc=0.5
+  - seaborn=0.6
+  - dask=0.7.5
+  - ipython=4.0.1
+  - sphinx=1.2.3  # pin to avoid https://github.com/sphinx-doc/sphinx/issues/1822

doc/examples/quick-overview.rst

@@ -139,7 +139,7 @@ index labels and dimensions. It looks a lot like a netCDF file:
 
 .. ipython:: python
 
-    ds = data.to_dataset()
+    ds = data.to_dataset(name='foo')
     ds
 
 You can do almost everything you can do with ``DataArray`` objects with

doc/indexing.rst

@@ -176,10 +176,10 @@ simultaneously, returning a new dataset:
 
 .. ipython:: python
 
-    ds = arr.to_dataset()
+    ds = arr.to_dataset(name='foo')
     ds.isel(space=[0], time=[0])
     ds.sel(time='2000-01-01')
-    ds2 = da.to_dataset()
+    ds2 = da.to_dataset(name='bar')
     ds2.isel_points(x=[0, 1, 6], y=[0, 1, 0], dim='points')
 
 Positional indexing on a dataset is not supported because the ordering of

readthedocs.yml

@@ -0,0 +1,5 @@
+conda:
+    file: doc/environment.yml
+python:
+   version: 2
+   setup_py_install: true