This is a repository to share the analysis of the home-credit-default-risk from kaggle.com (https://www.kaggle.com/c/home-credit-default-risk).

This repository was structured as below:

• 1-script

  This folder contain all script file that will be used to discover new insight from this project.

• 2-data

  This folder contain the data file used on analysis, all files was downloaded from "https://www.kaggle.com/c/home-credit-default-risk/data". Each new column on the files was create to improve analysis, all the new columns is available on file "HomeCredit_columns_description".

• 3-logs

  This folfer was created to store all logs o execution, this is userfull to identify the possible performance problem.

• 4-config

  This folder was created to store all configurable parameter that will be exposed. This directory will also contain environment settings that are required for all process analysis.

To make more easy to reproduce the result presente here, all python libraries used will be available at "4-config/config_linux_env.yml" and could be used to create a new conda env.

• Linux environment

    conda env create -f config_linux_env.yml
    conda source activate hcdr-project

After create a new conda enviroment it's necessary create a new ipkernel, juste follow the next step bellow to do that.

• Start conda enviroment

    source activate hcdr-project

• Install ipykernel on conda enviroment

    conda install ipykernel

• Create a new ipkernel.

    python -m ipykernel install --user --name hcdr-project --display-name hcdr-project