Getting-Started

This repository contains instructions to get started with Ansys Machine Learning and is intended to guide you through the workflow and common practices we encourage at Ansys. In the following sections, you can find instructions on how to prepare code in python, how to setup virtual environments or Docker containers and how to organize github repository.

Code Development in Python

A significant portion of machine learning development at Ansys takes place in Python.

1. In order to install Python, follow the instruction given at Install Python.
2. You can use any IDE you prefer. Most developers use PyCharm which provides user-friendly interface for code preparation and debugging.
3. We strongly encourage you to use virtual environments or Docker containers to compartmentalize your libraries/projects.
4. When you develop your code, please follow the PEP 8 coding style.
5. Testing: Test functions can help tremendously when multiple developers contribute to a single project. When a new developer adds new feature to an existing codebase, they will run the existing tests to ensure they have not broken any of the already implemented features. Then, they write additional unittests for the newly added features for future developers. A fairly popular way of writing test functions is to use the unittest library which is packaged with python. A light introduction of how to use unittest is provided in unittest introduction.
6. Documentation: There is a famous quote that "documentation is a gift to your future self". If you code is ready to be published consider adding consistent documentations. Please follow a standard scheme for documentation. You can read more about these schemes here. As an example, the area.py module in the example code folder follows a google docstring style. You can generate HTML or LaTex documentations using tools such as sphinx library.

Code Development in C++

TBA

Version Control

Version control is an essential part of any software engineering project.

1. At Ansys we use Git for version control. A short instruction on how to install and use Git for project development can be found at Git.
2. To share your code with others, interns can use our Ansys Research Github.
3. We strongly encourage you to follow the common project format for your machine learning projects.

Additional resources

(TO BE COMPLETED)

• If you're using Pytorch, have a look at Pytorch guidelines

• If you're using Tensorflow, have a look at Tensorflow guidelines

• If you're interested to see if you can run your code faster, have a look at faster code guidelines