One of the cool features that python has, is the ability to create a virtual environment similar to the virtual OS mechanism. The virtual environment is like a sandbox with its own site-packages and other libraries mimicking the python installed in the system. This kind of environment helps the developer/engineer to understand the deployment issues of the python library by installing and debugging the library dependencies. This step helps in identifying and fixing the dependency issues before deploying them into production environments. Also, as this entire stuff happens in an isolated virtual environment, he/she doesn't need to worry about the false results by OS level setup.
