DSLR

Data Science X Logistic Regression

Installation and dependencies

You need python, uv, make

git clone https://github.com/plagache/DSLR
cd DSLR
make setup

Usage

Mandatory part

# display information for all numerical features of a given dataset
make describe

# display respectively a histogram plot, a scatter plot and a pair plot
make histogram
make scatter
make pair

# train our model with gradient descent
make train

# generate a prediction file houses.csv
make predict

Bonus part

# launch a webserver that allow to display all the plot nicely
make debugweb

# launch a gradio server to tune our training easily
make gradio_interface

# implementation with scikit, for comparaison only
make scikit_logreg

# train with k-fold cross validation
make cross_validation