https://github.com/glenmiracle18/db_design_formative
no sql | mysql | convolutional neural network | fast-api
https://github.com/glenmiracle18/db_design_formative
Last synced: about 2 months ago
JSON representation
no sql | mysql | convolutional neural network | fast-api
- Host: GitHub
- URL: https://github.com/glenmiracle18/db_design_formative
- Owner: glenmiracle18
- Created: 2025-03-10T15:17:08.000Z (3 months ago)
- Default Branch: main
- Last Pushed: 2025-03-11T14:37:23.000Z (3 months ago)
- Last Synced: 2025-03-11T15:45:00.363Z (3 months ago)
- Language: Python
- Size: 55.1 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# DATABSE DESIGN FORMATIVE (PEER GROUP 2)
## Task 1: Database Design and Schema
Contributor: Glen Miracle
Responsibilties:
The following are the main things I was responsible for:
- Database Creation
- Entity Relation Diagram
- Database Schema design
- Seeding Scripts
- Creating Mongodb clusters and db's
- deploying databases
# Schema Files
```sql
-- Create Customers table
CREATE TABLE Customers (
customer_id VARCHAR(10) PRIMARY KEY, -- Assuming CustomerID is a string like "CUST123"
age INT NOT NULL CHECK (age >= 18),
gender INT NOT NULL CHECK (gender IN (0, 1)), -- 0 = Male, 1 = Female
income DECIMAL(10, 2) NOT NULL CHECK (income >= 0)
);
-- Create Visits table
CREATE TABLE Visits (
visit_id SERIAL PRIMARY KEY,
customer_id VARCHAR(10) REFERENCES Customers(customer_id) ON DELETE CASCADE,
visit_frequency VARCHAR(20) NOT NULL CHECK (visit_frequency IN ('Daily', 'Weekly', 'Monthly', 'Rarely')),
average_spend DECIMAL(6, 2) NOT NULL CHECK (average_spend >= 0),
preferred_cuisine VARCHAR(50) NOT NULL,
wait_time INT NOT NULL CHECK (wait_time >= 0),
service_rating INT NOT NULL CHECK (service_rating BETWEEN 1 AND 5),
food_rating INT NOT NULL CHECK (food_rating BETWEEN 1 AND 5),
ambiance_rating INT NOT NULL CHECK (ambiance_rating BETWEEN 1 AND 5)
);
-- Create Satisfaction table
CREATE TABLE Satisfaction (
satisfaction_id SERIAL PRIMARY KEY,
customer_id VARCHAR(10) REFERENCES Customers(customer_id) ON DELETE CASCADE,
visit_id INT REFERENCES Visits(visit_id) ON DELETE CASCADE,
satisfaction BOOLEAN NOT NULL, -- 0 = Not Satisfied, 1 = Satisfied
recorded_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
-- Stored Procedure: Insert a new customer visit and satisfaction record
CREATE OR REPLACE PROCEDURE add_customer_visit(
p_customer_id VARCHAR,
p_age INT,
p_gender INT,
p_income DECIMAL,
p_visit_frequency VARCHAR,
p_average_spend DECIMAL,
p_preferred_cuisine VARCHAR,
p_wait_time INT,
p_service_rating INT,
p_food_rating INT,
p_ambiance_rating INT,
p_satisfaction BOOLEAN
)
LANGUAGE plpgsql AS $$
DECLARE
v_visit_id INT;
BEGIN
-- Insert into Customers if not exists
INSERT INTO Customers (customer_id, age, gender, income)
VALUES (p_customer_id, p_age, p_gender, p_income)
ON CONFLICT (customer_id) DO NOTHING;
-- Insert into Visits
INSERT INTO Visits (customer_id, visit_frequency, average_spend, preferred_cuisine, wait_time, service_rating, food_rating, ambiance_rating)
VALUES (p_customer_id, p_visit_frequency, p_average_spend, p_preferred_cuisine, p_wait_time, p_service_rating, p_food_rating, p_ambiance_rating)
RETURNING visit_id INTO v_visit_id;
-- Insert into Satisfaction
INSERT INTO Satisfaction (customer_id, visit_id, satisfaction)
VALUES (p_customer_id, v_visit_id, p_satisfaction);
END;
$$;
-- Trigger: Log changes to Satisfaction table
CREATE TABLE Satisfaction_Log (
log_id SERIAL PRIMARY KEY,
satisfaction_id INT,
old_satisfaction BOOLEAN,
new_satisfaction BOOLEAN,
change_timestamp TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
CREATE OR REPLACE FUNCTION log_satisfaction_changes()
RETURNS TRIGGER AS $$
BEGIN
IF TG_OP = 'UPDATE' THEN
INSERT INTO Satisfaction_Log (satisfaction_id, old_satisfaction, new_satisfaction)
VALUES (OLD.satisfaction_id, OLD.satisfaction, NEW.satisfaction);
END IF;
RETURN NEW;
END;
$$ LANGUAGE plpgsql;
CREATE TRIGGER satisfaction_change_trigger
AFTER UPDATE ON Satisfaction
FOR EACH ROW
EXECUTE FUNCTION log_satisfaction_changes();
```
## ERD Diagram
## Task 2: APIs
Contributor: Jordan Nguepi
-created api for all CRUD operations on the 3 data models
---
`Data Models`
Customer, Visit, Satisfaction
api address:
`https://db-design-formative.onrender.com/api/'
available endpoint:
---
`api/customers/`
methods: LIST, CREATE, UPDATE, DELETE
---
`api/visits/`
methods: LIST, CREATE, UPDATE, DELETE
---
`api/satisfactions/`
methods: LIST, CREATE, UPDATE, DELETE
---
## **Task 3 - Fetching Data & Predicting Customer Satisfaction**
**Contributor:** Peter Johnson
### **Responsibilities**
In this project, I handled Task 3, which involved:
- Fetching the latest customer data from the API
- Preprocessing the data (categorical encoding, feature scaling)
- Loading and using a trained neural network model to predict customer satisfaction
### **Key Steps Implemented**
1. **API Integration**
- Connected to the Django API to retrieve the most recent customer entry
- Extracted the last entry
2. **Data Preprocessing**
- Standardized numerical features with Feature Scaling
3. **Model Deployment**
- Trained and used a neural network model (Best Model From Intro To ML)
- Loaded `trained_customer_satisfaction_model.keras` for predictions
4. **Database Logging**
- Logged predictions into the PostgreSQL database
### **Technologies Used**
- Python
- TensorFlow/Keras
- Django REST Framework
- PostgreSQL
- Scikit-Learn
### **Repository Structure**
- `db_directory/`→ Database Schema and tables
- `api_directory/` → Django API
- `ml-directory/` → ML model & and predictions
- `Task_3_PETER_JOHNSON.ipynb` → Task 3 implementation
- `trained_customer_satisfaction_model.keras` → Trained model