An example of how to create a simple self-developed Logistic Regression from scratch in Python
Shows how to use LogisticRegression.py
Contains the Logistic Regression using gradient descent
You can pull the repository the way it is. Running Main.py calls Logistic Regression with the iris dataset (with the first 100 samples). If you just want to use the Logistic Regression you have to optionally initialize it with the number of iterations and the learning rate and then call fit() with a Numpy Array X (m_samples, n_features) and a Numpy Array y (m_samples).
If you have an idea how to improve this Logistic Regression keeping it as simple as possible please fork it and make a PR. I'm not 100% sure if it is working correctly, because there is no 'clean' Logistic Regression in scikit-learn with which I can compare my results.