Merge pull request #58 from guillaumeeb/doc_correction

minor doc fixes

docs/examples.rst

@@ -15,7 +15,7 @@ PBS Deployments
 
    cluster = PBSCluster(queue='regular',
                         project='DaskOnPBS',
-                        local_directory=os.getenv('TMPDIR', '/tmp'),
+                        local_directory='$TMPDIR',
                         threads=4,
                         processes=6,
                         memory='16GB',

docs/index.rst

@@ -8,8 +8,8 @@ PBS, Slurm, and SGE.*
 Motivation
 ----------
 
-1. While ``dask.distributed`` offers a flexible distributed parallel computing
-   Python, it is not always easy to deploy on systems that use job queuing
+1. While ``dask.distributed`` offers a flexible library for distributed computing
+   in Python, it is not always easy to deploy on systems that use job queuing
    systems. Dask-jobqueue provides a Pythonic interface for deploying and
    managing Dask clusters.
 2. In practice, deploying distributed requires customization, both for the

docs/install.rst

@@ -10,7 +10,8 @@ Pip can be used to install both dask-jobqueue and its dependencies (e.g. dask,
 distributed,  NumPy, Pandas, and so on that are necessary for different
 workloads).::
 
-   pip install "dask_jobqueue"    # Install everything
+   pip install "dask_jobqueue"    # Install everything from last released version
+
 
 Install from Source
 -------------------
@@ -26,11 +27,16 @@ or use ``pip`` locally if you want to install all dependencies as well::
 
     pip install -e .
 
+You can also install directly from git master branch::
+
+    pip install git+https://github.com/dask/dask-jobqueue
+
 
 Test
 ----
 
-Test dask with ``py.test``::
+Test dask-jobqueue with ``py.test``::
 
-    cd dask
+    git clone https://github.com/dask/dask-jobqueue.git
+    cd dask-jobqueue
     py.test dask_jobqueue