Ajout de la documentation et de la configuration Backstage.io

catalog-info.yaml

@@ -0,0 +1,18 @@
+apiVersion: backstage.io/v1alpha1
+kind: Component
+metadata:
+  name: loan-approval-predictor
+  description: A machine learning service for loan approval prediction
+  annotations:
+    github.com/project-slug: loan-approval-predictor
+    backstage.io/techdocs-ref: dir:./docs
+  tags:
+    - fastapi
+    - python
+    - machine-learning
+    - finance
+spec:
+  type: service
+  lifecycle: experimental
+  owner: finance-team
+  system: ml-services

docs/api-reference.md

@@ -0,0 +1,44 @@
+# API Reference
+
+## Endpoints
+
+### GET /
+
+Returns the home page HTML.
+
+### GET /predict
+
+Returns the prediction form HTML.
+
+### POST /predict
+
+Predicts loan approval status based on the provided applicant information.
+
+**Request Parameters**
+
+| Parameter | Type | Description |
+|-----------|------|-------------|
+| Gender | int | Gender of the applicant (0 for Female, 1 for Male) |
+| Married | int | Marital status (0 for No, 1 for Yes) |
+| Dependents | int | Number of dependents (0, 1, 2, or 3+) |
+| Education | int | Education level (0 for Not Graduate, 1 for Graduate) |
+| Self_Employed | int | Self-employment status (0 for No, 1 for Yes) |
+| ApplicantIncome | float | Applicant's income |
+| CoapplicantIncome | float | Co-applicant's income |
+| LoanAmount | float | Loan amount in thousands |
+| Loan_Amount_Term | int | Term of loan in months |
+| Credit_History | int | Credit history (0 for No, 1 for Yes) |
+| Property_Area | int | Area of property (0 for Rural, 1 for Semiurban, 2 for Urban) |
+
+**Response**
+
+The response is an HTML page with the prediction result:
+- "Approved" if the loan is predicted to be approved
+- "Not Approved" if the loan is predicted to be rejected
+
+## Error Handling
+
+The API uses standard HTTP status codes to indicate the success or failure of requests.
+
+- 200: Success
+- 500: Server error

docs/getting-started.md

@@ -0,0 +1,58 @@
+# Getting Started
+
+## Prerequisites
+
+- Docker
+- Python 3.7+
+
+## Installation
+
+1. Clone the repository
+2. Build the Docker image:
+
+```bash
+docker build -t loan-approval-predictor .
+```
+
+3. Run the container:
+
+```bash
+docker run -p 8000:8000 loan-approval-predictor
+```
+
+## Development
+
+To set up a development environment:
+
+1. Create a virtual environment:
+
+```bash
+python -m venv venv
+source venv/bin/activate  # On Windows: venv\Scripts\activate
+```
+
+2. Install dependencies:
+
+```bash
+pip install -r requirements.txt
+```
+
+3. Run the application:
+
+```bash
+uvicorn main:app --reload
+```
+
+The application will be available at http://localhost:8000.
+
+## Testing
+
+Run the tests using:
+
+```bash
+pytest test_app.py
+```
+
+## Deployment
+
+The application includes a GitHub Actions workflow for CI/CD in the `.github/workflows/deploy.yml` file.

docs/index.md

@@ -0,0 +1,27 @@
+# Loan Approval Predictor
+
+## Overview
+
+This is a machine learning service that predicts loan approval based on applicant information. The service uses a pre-trained random forest model and provides both a web interface and API endpoints for predictions.
+
+## Features
+
+- Loan approval prediction based on applicant data
+- Web interface for easy interaction
+- RESTful API for programmatic access
+- Containerized with Docker
+- CI/CD pipeline with GitHub Actions
+
+## Architecture
+
+The service is built using:
+- FastAPI for the web framework
+- Scikit-learn for the machine learning model (Random Forest)
+- Docker for containerization
+- GitHub Actions for CI/CD
+
+## Documentation
+
+- [API Reference](api-reference.md)
+- [Model Information](model-info.md)
+- [Getting Started](getting-started.md)

docs/model-info.md

@@ -0,0 +1,43 @@
+# Model Information
+
+## Overview
+
+The Loan Approval Predictor uses a Random Forest machine learning model trained on loan application data to predict whether a loan application will be approved or rejected.
+
+## Model Details
+
+- **Model Type**: Random Forest Classifier
+- **Features Used**:
+  - Gender
+  - Married
+  - Dependents
+  - Education
+  - Self_Employed
+  - ApplicantIncome
+  - CoapplicantIncome
+  - LoanAmount
+  - Loan_Amount_Term
+  - Credit_History
+  - Property_Area
+- **Output**: Binary prediction (Approved/Not Approved)
+- **Model File**: `random_forest_model.pkl`
+- **Scaler File**: `scaler.pkl`
+
+## Dataset
+
+The model was trained on the Loan Approval Prediction dataset, which contains historical loan application data with features about applicants and whether their loans were approved.
+
+## Feature Importance
+
+The most important features for loan approval prediction typically include:
+1. Credit History
+2. Loan Amount
+3. Applicant Income
+4. Property Area
+5. Loan Amount Term
+
+## Limitations
+
+- The model is trained on historical data and may not capture recent changes in lending policies
+- The model assumes that the input data follows the same distribution as the training data
+- Extreme outlier values may lead to unreliable predictions

mkdocs.yml

@@ -0,0 +1,12 @@
+site_name: 'Loan Approval Predictor'
+nav:
+  - Home: index.md
+  - API Reference: api-reference.md
+  - Model Information: model-info.md
+  - Getting Started: getting-started.md
+plugins:
+  - techdocs-core
+markdown_extensions:
+  - admonition
+  - pymdownx.highlight
+  - pymdownx.superfences