Installation#

Installing with pip#

DoWhy support Python 3.6+. To install, you can use pip or conda.

Latest Release

Install the latest release using pip.

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.

pip install -e .

Requirements

If you face any problems, try installing dependencies manually.

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).

Note

Installing pygraphviz can cause problems on some platforms. One way that works for most Linux distributions is to first install graphviz and then pygraphviz as shown below. Otherwise, please consult the documentation of pygraphviz.

sudo apt install graphviz libgraphviz-dev graphviz-dev pkg-config
pip install --global-option=build_ext \
--global-option="-I/usr/local/include/graphviz/" \
--global-option="-L/usr/local/lib/graphviz" pygraphviz

Installing with Conda#

Install the latest release using conda.

conda install -c conda-forge dowhy

If you face “Solving environment” problems with conda, then try conda update --all and then install dowhy. If that does not work, then use conda config --set channel_priority false and try to install again. If the problem persists, please add your issue here.

Installing on Azure Machine Learning#

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

Open a new terminal window - it will open pointing to your user folder
Create a new folder (if you wish - this is not really necessary):
```
mkdir pywhy
```
To be really pedantic, ensure it is fully ‘activated’:
```
chmod 777 pywhy
```
Get the full path by:
```
cd pywhy
pwd
```
Copy that path you will need it later.

Clone the repository:

git clone https://github.com/py-why/dowhy

Now open a python notebook and create a new python code cell. Type:
```
%pip install -e <path from step d.>
```
To test the installation:
```
import dowhy
```

This should run with no errors.