- Exploratory data analysis (see eda.ipynb) presented in a streamlit dashboard
- Check the dashboard in the following link:
https://dashboard-deployment-deliverable-jedha.streamlit.app
- Car rental price predictor
Users are needed to provide the following features on which the predicting model was trained on:
-
brand (string): The brand of the car (e.g., "Toyota").
-
mileage (int): The mileage of the car in kilometers (e.g., 35000).
-
engine_power (int): The power of the car's engine in horsepower (e.g., 110).
-
type_of_fuel (string): The type of fuel the car uses (e.g., "diesel").
-
color (string): The color of the car (e.g., "black").
-
type_of_car (string): The category of the car (e.g., "suv").
-
private_parking_available (bool): Whether private parking is available (e.g., "yes" or "no").
-
gps (bool): Whether the car has a GPS (e.g., "yes" or "no").
-
air_conditioning (bool): Whether the car has air conditioning (e.g., "yes" or "no").
-
automatic_car (bool): Whether the car is automatic (e.g., "yes" or "no").
-
connect_check_in (bool): Whether the car supports GetAround Connect for remote check-in (e.g., "yes" or "no").
-
speed_regulator (bool): Whether the car has a speed regulator (e.g., "yes" or "no").
-
winter_tires (bool): Whether the car has winter tires (e.g., "yes" or "no").
-
Errors in the request are treated with the corresponding correction in the respose body.
This is the link of the API : https://get-around-app-2b184b2f74b6.herokuapp.com/docs
response = requests.post("https://get-around-app-2b184b2f74b6.herokuapp.com/predict_price", json={ "input": [['Toyota', 140411, 100, 'diesel', 'black', 'convertible', 'yes', 'yes', 'no', 'no', 'yes', 'yes', 'yes']] }) print(response.json())
Or on terminal (Mac) in the following way:
curl -i -H "Content-Type: application/json" -X POST -d '{"input": [["Toyota", 140411, 100, "diesel", "black", "convertible", "yes", "yes", "no", "no", "yes", "yes", "yes"]]}' https://get-around-app-2b184b2f74b6.herokuapp.com/predict_price
The API /predict_from_browser endpoint, allows you predict the price of a car rental from your browser. It takes the above-mentioned information as input and returns the predicted rental price of the car.
-The output is a JSON object with the predicted price. Example: {"predicted_price": 49.99}