Added a section about binary builds.

Author: StuntsPT

docs/binaries.md

@@ -0,0 +1,39 @@
+# Binary building
+
+For your conveninence, we have pre-build binaries of STRUCTURE and
+fastStructure. They are provided with the package under
+`structure_threader/bins/$platform/`.
+Here is how they were built.
+
+## The build system
+
+### GNU/Linux binaries
+Binaries were built on a machine with an Intel Xeon E5-2609 0 @ 2.40GHz CPU.
+The OS under which the binaries were built is Ubuntu 12.04 64bit. This "old"
+OS was used since linux systems have backwards, but not forwards compatibility.
+This means that binaries built on older systems will run on newer systems, but
+the opposite may not be true.
+
+### OSX binaries
+We have not built any OSX binaries yet. We will update this section once we do.
+
+
+## STRUCTURE
+
+STRUCTURE is relatively simple to build. Source code can be obtained from the
+ [STRUCTURE website](http://pritchardlab.stanford.edu/structure_software/release_versions/v2.3.4/structure_kernel_source.tar.gz). To build the binary, we used our helper
+ script "install_structure.sh".
+
+
+ ## fastStructure
+
+fastStructure requires a more involved process to build as it requires many
+dependencies. It can be obtained from
+[it's own github repository](https://github.com/rajanil/fastStructure).
+Although fastStructure is written in python 2, it uses compiled code, by making
+use of `cython`.
+To build the binary, we have insatlled fastStructure using our helper script
+"install_faststructure.sh" and then we have used
+[pyinstaller](http://www.pyinstaller.org/) to turn it into a binary. The used
+"specfile" can be found [here](https://github.com/StuntsPT/Structure_threader/tree/master/helper_scripts/structure.spec) which contains all the required information to reproduce the
+build.

docs/index.md

@@ -16,6 +16,7 @@ To run "fastChooseK.py" (fastStructure wrapper only), numpy is also required.
 
 ## Contents
 * [Installation & dependencies](install.md)
+    * [Binary building](binaries.md)
 * [Usage](usage.md)
 * [Output](output.md)
 * [Test Data](test_data.md)