Skip to main content

team21

Continuous Integration Test Coverage for Milestone 2 test AD for Milestone 2

Brief Introduction

funAD is a PyPi-distributed package that executes forward-mode of automatic differentiation, enabling users to solve functional derivatives with high computational efficiency and machine precision.

This project/package is the fruit of Harvard CS107/AC207 class final project in 2022 Fall. Our package utilize forward mode and dual number. Additionally, we also allow users to define their own seeds vector for the Jacobian Matrix and the option to calculate local maxima and minima through gradient descent.

To install, run the following command in your terminal

pip install funAD

For details instruction on how to use this package, please follow the steps in the usage page.

Group Number:

Group 21

Group Members:

Zhecheng Yao zhechengyao@g.harvard.edu

Hanlin Zhu hzhu@g.harvard.edu

Xu Tang xutang@g.harvard.edu

Tiantong Li tiantongli@g.harvard.edu

Harvard University, Fall 2022

Broader Impact

funAD is a useful tool to reduce the computational burden by reducing the amount of manual computation and providing accurate results. Our package supports the forward mode of automatic differentiation, as well as gradient descent optimizer. funAD enables users to compute derivatives of complex functions with machine precision and high computational efficiency. Additional optimization function is implemented to generate local minima and maxima for the user. These functions mentioned above are applicable to many research areas. For example, computational biologists often need to perform the above calculations in order to estimate population growth and peaks for the target species. Optimizers such as gradient descent help minimize loss functions, which play a significant role in the development of neural networks, computer vision, and many statistical models. With the assistance of optimization functions and its computationally efficient nature, funAD can be very useful in training machine learning models and neural networks.This may also benefit economists and financial researchers who can minimize cost functions and maximize profit functions in the context of industrial, business and investment analysis to generate the best returns for their companies.

Unlike some models trained based on datasets, our package is essentially a specified calculator that follows mathematical expressions and rules. The package itself is not biased in any way. However, users should be careful not to use it in a discriminative way such as training machine learning models on biased datasets. Users need to discipline themselves not to use our package to bypass the calculation step when they need to find derivative results manually as a math assignment. Although the code is restricted to code.harvard.edu for grading purposes, our package is uploaded to PyPI, which is free to download and relatively easy to implement. Our email addresses are listed on PyPI's funAD package download page. Users can easily contact us with any feedback and suggestions.

Metadata

Release files for funAD 1.0.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for funAD 1.0.1
File Size Uploaded Provenance Actions
funAD-1.0.1.tar.gz 11.6 kB Without provenance Details for funAD-1.0.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for funAD 1.0.1
File Interpreter ABI Platform Actions
funAD-1.0.1-py3-none-any.whl Python 3 none any Details for funAD-1.0.1-py3-none-any.whl

Total release size: 23.0 kB

Release files / funAD-1.0.1.tar.gz

Download URL funAD-1.0.1.tar.gz
Size 11.6 kB
Tags Source
SHA-256 checksum
How to use checksums
ab69dbb421d38b4c679901de86227857dbc3862d685aae35f24d734f304742f2
BLAKE2b-256 checksum
How to use checksums
be3b43a45e5356133ee9bda20d726808bee786d84f9ea5985fe6d26d246be484
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/3.8.0 colorama/0.4.4 importlib-metadata/4.6.4 keyring/23.5.0 pkginfo/1.8.2 readme-renderer/34.0 requests-toolbelt/0.9.1 requests/2.25.1 rfc3986/1.5.0 tqdm/4.57.0 urllib3/1.26.5 CPython/3.10.6

Release files / funAD-1.0.1-py3-none-any.whl

Download URL funAD-1.0.1-py3-none-any.whl
Size 11.4 kB
Tags Python 3
SHA-256 checksum
How to use checksums
19c9863e9d40c904ef65397696cbe8bb7be3b99e676efa3aeb1f969d31746f76
BLAKE2b-256 checksum
How to use checksums
11c252d5689e04054a65580ad81cda652630aad7af92c1e07ad99945c261b0ac
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/3.8.0 colorama/0.4.4 importlib-metadata/4.6.4 keyring/23.5.0 pkginfo/1.8.2 readme-renderer/34.0 requests-toolbelt/0.9.1 requests/2.25.1 rfc3986/1.5.0 tqdm/4.57.0 urllib3/1.26.5 CPython/3.10.6

Security for funAD 1.0.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page