Skip to content

Python Flask-2.md

Latest commit

 

History

History
215 lines (136 loc) · 2.96 KB

Python Flask-2.md

File metadata and controls

215 lines (136 loc) · 2.96 KB

Building an API using Flask (For ML Projects)

🧠 Mental Model

“An API is just a URL that returns data, not HTML.”

Example:

/predict → returns result

No page. No UI. Only data.

1️⃣ Install Flask

pip install flask

“Flask is lightweight. Perfect for ML.”

2️⃣ Minimal Flask API Structure

Tell them to create:

ml_api/
│── app.py

“For APIs, we don’t need templates.”

3️⃣ Create Your First API Endpoint

app.py

from flask import Flask, jsonify

app = Flask(__name__)

@app.route("/api/hello")
def hello_api():
    return jsonify({
        "message": "Hello from Flask API"
    })

if __name__ == "__main__":
    app.run(debug=True)

Say while explaining:

  • jsonify → converts Python dict to JSON

  • Browser shows JSON, not HTML

4️⃣ Test the API (Important Habit)

Run:

python app.py

Open browser:

http://127.0.0.1:5000/api/hello

“If this works,
you have officially built an API.”

Let that moment land.

5️⃣ Accept Input Data (API with Parameters)

“ML models need input.
APIs receive input as JSON.”

Update API

from flask import request

@app.route("/api/square", methods=["POST"])
def square_number():
    data = request.get_json()
    number = data["number"]

    result = number * number

    return jsonify({
        "input": number,
        "output": result
    })

Say slowly:

  • POST → sending data

  • request.get_json() → read input

  • Return JSON result

6️⃣ How to Test This API (No Frontend Yet)

“APIs are tested using tools, not browser typing.”

Using Postman / Thunder Client / curl

Example JSON:

{
  "number": 5
}

Response:

{
  "input": 5,
  "output": 25
}

“This is how frontend talks to backend.”

7️⃣ Connect This to ML (Key Moment)

Now relate it to ML, verbally:

“Replace this logic with your ML model prediction.”

Example explanation:

prediction = model.predict(input_data)

“API does not care how complex your model is.
It only sends input and receives output.”

This clicks instantly.

8️⃣ Simulating an ML Prediction (Beginner Friendly)

Fake ML Logic (For Understanding)

@app.route("/api/predict", methods=["POST"])
def predict():
    data = request.get_json()
    marks = data["marks"]

    if marks >= 50:
        result = "Pass"
    else:
        result = "Fail"

    return jsonify({
        "marks": marks,
        "prediction": result
    })

“Today this is logic.
Tomorrow this will be your ML model.”

9️⃣ Where UI Fits (Very Simple Explanation)

“UI is optional.
API is mandatory.”

Explain flow:

HTML / React / Mobile App
            ↓
         Flask API
            ↓
         ML Model
            ↓
         JSON Response

Don’t build UI yet.