This project is aimed at fulfilling the datascience exercice provided by Xebia
*To build the docker type from within the repository :
docker build -t mower_market:dev:dev -f Dockerfile.dev .
*Once built use to start the docker as in stance for Jyputer :
docker run -it -p 8889:8888 -p 5000:5000 -p 6006:6006 --rm -v //c/Users/thilegall/Desktop/Projets/Xebia/mower_market:/root mower_market:dev jupyter notebook --allow-root
*Finally, you can access your Jupyter notebook at (in any browser) :
localhost:8889
#Filesystem | (dockerfilesystem)
(/root)
|-mower_mark
|
|-API -> File containing a set of function for datascience purpose
|-crunching.py -> data crunching functions
|-features.py -> feature_engineering functions
|-finalistaion.py -> specific function linked to a project (fitting stastical distibution, etc...)
|-model_application.py -> functions that apply the selected model
|-model_testing.py -> functions used to tes, evaluate and register models
|-data -> File containing the data regarding the project
|-crunching -> scripts for data crunching (empty)
|-DF -> output data
|-features -> scripts for feature engineering (empty)
|-finalisation -> scripts for finalisation (empty)
|-model -> models
|-mower_market_datasets -> input data
|-ref -> File containing a reference table with the model and their metadata (performance, associated scripts, etc...) (empty)
|-scaler -> scaler of the data (if the model is not a scikitlearn model purpose) (empty)
|-LOG -> File where the logs will be written
|-notebook -> File containing the notebooks
|-data_explo.ipnb -> the exploration and cleaning notebook
|-tests_scripts_crunching.ipnb -> test of crunching fucntion called as API + data modelling
|-test_main.ipnb -> tests of the main function to call it as a script
|-source -> File containing the script rnning the project
|-main.py -> script used to run the main
|-utils -> File containing utility functions (empty)