Skip to content

A Python library called regresspy may be used to do straightforward regressions using the gradient descent approach. The library, however, is just halfway finished.

License

Notifications You must be signed in to change notification settings

Mainul98/regresspy-mainul

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

regresspy

regresspy is a python module for carrying out simple regressions using gradient descent algorithm. However, the library is half complete. Your task is to fill in the necessary codes. The tasks of the project is given below:

  • Fork this repository.
  • Create a new branch using your username.
  • Fill in the .gitignore file by creating one for Python from .gitignore.io.
  • Commit frequently, usually everytime after changing or completing a function. Make the commit messages clear and complete. You may read this blog to learn about writing good commit messages.
  • Write the codes for the mean absolute, sum of squared, mean squared, and root mean squared errors in the loss.py file.
  • Complete the test_loss.py by filling in the codes to test above functions. An example for testing mae is already given. You may read this blog or any other to learn about pytest.
  • After that, go to the gradient_descent.py file. It contains two functions named forward for computing forward propagation and backward for computing the gradients. The second one is already done for you. All you have to do is fill in the codes for forward propagation.
  • Then go to the main.py file where the actual training will take place. The first task is to complete the _initialize_weights function. Then complete the _train function. Finally, complete the predict and score function.
  • Now, go the model.py file in the test folder. Here, you will perform two regressions. First one will be carried out using the SGDRegressor from the sklearn library. Second one will be carried out using your own codes. Compare the results and write in the README.md file (You can delete this README file).
  • Fill in the requirements.txt file according to the libraries you have used in your codes.
  • Fill-in the setup.py file to install your code as a library. You may read this blog or the Python documentation to learn how to prepare this file.
  • Give an appropriate license to your library in the LICENSE file. You may use choosealicense.com to select the best one according to your case.
  • Try to make your code look beautiful. You can add optional documentation about how to use your library in the README file. Look for other repositories and see how they have documented.
  • Finally, email me the link of your repository.

About

A Python library called regresspy may be used to do straightforward regressions using the gradient descent approach. The library, however, is just halfway finished.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages