Software Organization¶
Directory Structure¶
gradpy\
gradpy\
tests\
__init__.py
test_autodiff.py
__init__.py
autodiff.py
math.py
setup.py
requirements.txt
LICENSE
README.md
docs\
milestone1.ipynb
examples\
introductory_demo.ipynb
special_functions_demo.ipynb
newton_multivar_demo.ipynb
newton_multivar.py
README.md
.travis.yml
setup.cfg
Modules¶
- gradpy/autodiff.py: defines the core structure of automatic differentiation, including attributes for accessing values and gradients, and methods for operator overloading. This is the main module users will import.
- gradpy/math.py: defines elementary functions (sin, exp) to be imported by the user and used when defining a function.
- gradpy/tests/test_autodiff.py: A module of test functions used in validating our implementation.
Examples¶
- docs/examples/introductory_demo.ipynb: Introductory examples of using the
gradpypackage, covering single-variable, multi-variable, and vector-valued functions. - docs/examples/special_functions_demo.ipynb: How to import and use functions such as
sinandexp. - docs/examples/newton_multivar_demo.ipynb: An application of
gradpyin root-finding using Newton's method.
Tests¶
Our test suite is maintained along with our package on GitHub, with Travis CI and Coveralls for continuous integration. Instructions for running the test suite after successfully installing gradpy can be found in the Installation section.