GitLocker: The Coding Marketplace

Description:

predictionmodelmanifoldailearning 1.0.0

Packaging the ML Model of Classification
Problem Statement

Company wants to automate the loan eligibility process based on customer detail provided while filling online application form.
It is a classification problem where we have to predict whether a loan would be approved or not.

Data
The data corresponds to a set of financial requests associated with individuals.

Variables
Description

Loan_ID
Unique Loan ID

Gender
Male/ Female

Married
Applicant married (Y/N)

Dependents
Number of dependents

Education
Applicant Education (Graduate/ Under Graduate)

Self_Employed
Self employed (Y/N)

ApplicantIncome
Applicant income

CoapplicantIncome
Coapplicant income

LoanAmount
Loan amount in thousands

Loan_Amount_Term
Term of loan in months

Credit_History
credit history meets guidelines

Property_Area
Urban/ Semi Urban/ Rural

Loan_Status
Loan approved (Y/N)

Source: Kaggle
Running Locally
Add PYTHONPATH variable for ~/.bash_profile for MacOS
</code></pre>
<h2>Virtual Environment</h2>
<p>Install virtualenv</p>
<pre lang="python"><code>python3 -m pip install virtualenv

Check version
virtualenv --version

Create virtual environment
virtualenv ml_package

Activate virtual environment
For Linux/Mac
source ml_package/bin/activate

For Windows
ml_package\Scripts\activate

Deactivate virtual environment
deactivate

Directory structure
prediction_model

├── MANIFEST.in
├── prediction_model
│   ├── config
│   │   ├── config.py
│   │   └── __init__.py
│   ├── datasets
│   │   ├── __init__.py
│   │   ├── test.csv
│   │   └── train.csv
│   ├── __init__.py
│   ├── pipeline.py
│   ├── predict.py
│   ├── processing
│   │   ├── data_handling.py
│   │   ├── __init__.py
│   │   └── preprocessing.py
│   ├── trained_models
│   │   ├── classification.pkl
│   │   └── __init__.py
│   ├── training_pipeline.py
│   └── VERSION
├── README.md
├── requirements.txt
├── setup.py
└── tests
├── pytest.ini
└── test_prediction.py

Build the Package

Goto Project directory and install dependencies
pip install -r requirements.txt

Create Pickle file after training:
python prediction_model/training_pipeline.py

Create source distribution and wheel
python setup.py sdist bdist_wheel

Installation of Package
Go to project directory where setup.py file is located

To install it in editable or developer mode

pip install -e .

. refers to current directory
-e refers to --editable mode

Normal installation

pip install .

. refers to current directory

Also can be installed from git as well after pushing to github

pip install git+https://github.com/manifoldailearning/prediction_model.git

Testing the Package Working

Remove the PYTHONPATH from environment variables
Goto a separate location which is outside of package directory
Create a new virual environment using the commands mentioned above & activate it
Before installing, test whether you are able to import the package of prediction_model - (you should not be able to do it)
Now in the new environment install the package from github
pip install git+https://github.com/manifoldailearning/prediction_model.git
Now try importing the prediction_model, you should be able to do it successfully
Extras : Run training pipeline using the package, and also conduct the test