Using PySpark Titanic Survival classifier is deployed using flask web app and expose a rest end point.

file information -

1. server.py - which contain get and post reuest function, in get request training a model and saving it, in post request it load model and reponse output on requested data.
2. preprocess_file.py - which contain all preprocess function and transformation steps.
3. train_model.py - it contain model training and saving code.

Step 1 - Clone this repo in your system. Step 2 - go to flask_app_deployment folder and run command - export FLASK_APP=server Step 3 - run command - flask run step 4 - open any rest client app ex. postman, ARC etc.

step 5 - hit url http://localhost:5000/ - it will start training and saving of model step 6 - hit url http://localhost:5000/predict - this is post request so you need to pass data in json format.

Data request in json format for http://localhost:5000/predict

[{ "PassengerId":894, "Pclass":1, "Name":"Kelly, Mrs. James", "Sex":"female", "Age":22, "SibSp":1, "Parch":1, "Ticket":"330911", "Fare":21, "Cabin":1, "Embarked":"S" }]