move install instructions to top of README

README.md

@@ -2,12 +2,26 @@
 Maximum likelihood and variational Bayes estimation for
 mixtures of regular Brownian motions
 
+## Install from source
+
+Clone the repo, then
+```
+cd emdiff
+pip install -e .
+```
+
+Then make sure you can run the example:
+```
+cd samples
+python sample_script.py
+```
+
 ## What does it do?
 
 `emdiff` is a Python analysis tool for trajectories generated by 
 stroboscopic single particle tracking (SPT) experiments.
 
-It takes a set of trajectories in CSV format, then returns the 
+It takes a set of trajectories as a `pandas.DataFrame`, then returns the 
 estimated fraction of molecules in distinct states and the 
 diffusion coefficient corresponding to each state. 
 
@@ -86,19 +100,6 @@ part of the standard [Anaconda](https://www.anaconda.com/products/individual) di
 [`matplotlib_scalebar`](https://pypi.org/project/matplotlib-scalebar/), 
 the plots will have scalebars. So that's recommended too.
 
-## Install
-
-1. Clone the repository: `git clone https://github.com/alecheckert/emdiff.git`. 
-
-2. Navigate to the top-level `emdiff` directory.
-
-3. From a `conda` environment with the dependencies above, run `python setup.py develop`. 
-
-`emdiff` is in active development. The `develop` option 
-will track changes in the source files as new versions become available.
-To pull new versions, navigate to the `emdiff` repository and use
-`git pull`. 
-
 ## Expected input
 
 Both `emdiff` and `vbdiff` take trajectories as a `pandas.DataFrame`. Each row of