docs overview

Author: Ehsan Totoni

docs/index.rst

@@ -3,18 +3,12 @@
    You can adapt this file completely to your liking, but it should at least
    contain the root `toctree` directive.
 
-Welcome to HPAT's documentation!
+HPAT documentation
 ================================
 
 .. toctree::
    :maxdepth: 2
    :caption: Contents:
 
-
-
-Indices and tables
-==================
-
-* :ref:`genindex`
-* :ref:`modindex`
-* :ref:`search`
+   source/overview
+   source/supported

docs/source/overview.rst

@@ -0,0 +1,33 @@
+
+Overview
+========
+
+High Performance Analytics Toolkit (HPAT) is a big data analytics and machine
+learning framework that provides Python's ease of use but is extremly fast.
+
+HPAT scales analytics programs in python to cluster/cloud environments
+automatically, requiring only minimal code changes. Here is a logistic
+regression program using HPAT::
+
+    @hpat.jit
+    def logistic_regression(iterations):
+        f = h5py.File("lr.hdf5", "r")
+        X = f['points'][:]
+        Y = f['responses'][:]
+        D = X.shape[1]
+        w = np.random.ranf(D)
+        t1 = time.time()
+        for i in range(iterations):
+            z = ((1.0 / (1.0 + np.exp(-Y * np.dot(X, w))) - 1.0) * Y)
+            w -= np.dot(z, X)
+        return w
+
+This code runs on cluster and cloud environments using a simple command like
+`mpirun -n 1024 python logistic_regression.py`.
+
+HPAT compiles a :ref:`subset of Python <supported>` to efficient native parallel code
+(with `MPI <https://en.wikipedia.org/wiki/Message_Passing_Interface>`_).
+This is in contrast to other frameworks such as Apache Spark which are
+master-executor libraries. Hence, HPAT is typically 100x or more faster.
+HPAT is built on top of `Numba <https://github.com/numba/numba>`_
+and `LLVM <https://llvm.org/>`_ compilers.

docs/source/supported.rst

@@ -0,0 +1,6 @@
+.. _supported:
+
+Supported Python Programs
+=========================
+
+The `@hpat.jit` decorator...