installation in Azure ML guide (#473)

* follow up on open issue on the same topic

One MarkDown file describing the neuances of installing within AzureML

* Moved file to the docs folder, added pip/conda installation instructions, and converted to rst

Signed-off-by: Amit Sharma <amit_sharma@live.com>

Co-authored-by: Amit Sharma <amit_sharma@live.com>

CONTRIBUTING.md

@@ -60,4 +60,4 @@ problems
 * Helping update the documentation for DoWhy
 
 If you would like to contribute, you can raise a pull request. If you have
-questions before contributing, you can start by opening an issue on Github. 
+questions before contributing, you can start by opening an issue on Github.

docs/source/getting_started/install.rst

@@ -1,2 +1,110 @@
-Installation
-============
+============
+Installation
+============
+
+Installing with pip
+-------------------
+
+DoWhy support Python 3.6+. To install, you can use pip or conda. 
+
+**Latest Release**
+
+Install the latest `release <https://pypi.org/project/dowhy/>`__ using pip.
+
+.. code:: shell
+   
+   pip install dowhy
+   
+**Development Version**
+
+If you prefer the latest dev version, clone this repository and run the following command from the top-most folder of
+the repository.
+
+.. code:: shell
+    
+    pip install -e .
+
+**Requirements**
+
+If you face any problems, try installing dependencies manually.
+
+.. code:: shell
+    
+    pip install -r requirements.txt
+
+Optionally, if you wish to input graphs in the dot format, then install pydot (or pygraphviz).
+
+
+For better-looking graphs, you can optionally install pygraphviz. To proceed,
+first install graphviz and then pygraphviz (on Ubuntu and Ubuntu WSL).
+
+.. code:: shell
+
+    sudo apt install graphviz libgraphviz-dev graphviz-dev pkg-config
+    ## from https://github.com/pygraphviz/pygraphviz/issues/71
+    pip install pygraphviz --install-option="--include-path=/usr/include/graphviz" \
+    --install-option="--library-path=/usr/lib/graphviz/"
+
+Installing with Conda
+---------------------
+
+Install the latest `release <https://anaconda.org/conda-forge/dowhy>`__ using conda.
+
+.. code:: shell
+
+   conda install -c conda-forge dowhy
+
+If you face "Solving environment" problems with conda, then try :code:`conda update --all` and then install dowhy. If that does not work, then use :code:`conda config --set channel_priority false` and try to install again. If the problem persists, please add your issue `here <https://github.com/microsoft/dowhy/issues/197>`_.
+
+
+Installing on Azure Machine Learning
+------------------------------------
+
+Eli Y. Kling {https://www.linkedin.com/in/elikling/}
+
+In Azure Machine Learning it is not that straight forward to identify in the terminal window the python (Conda) envornoments used by the notebook. Thus, it is easier to run shell commands from within the notebook. The secret is NOT to use the ! magic but the %.
+
+**Getting the latest release**
+
+In an new python code cell type::
+
+    %pip install dowhy
+
+Or::
+
+    %pip install --force-reinstall --no-cache-dir dowhy
+
+**Getting the dev version**
+
+a. Open a new terminal window - it will open pointing to your user folder
+
+b. Create a new folder (if you wish - this is not really necessary)::
+
+    mkdir pywhy   
+
+c. To be really pedantic, ensure it is fully 'activated'::
+
+    chmod 777 pywhy
+
+d. Get the full path by::
+
+    cd pywhy
+    pwd
+
+e. Copy that path you will need it later.
+
+f. Clone the repository::
+
+    git clone https://github.com/py-why/dowhy
+
+g. Now open a python notebook and create a new python code cell. Type::
+
+    %pip install -e <path from step d.>
+
+h. To test the installation::
+
+    import dowhy
+    
+
+This should run with no errors.
+