ExtroPlanner

API Web-application to aid in planing outdoor events/activities and monitor the current environment.

Members:

Name	ID	Department	Faculty
Khunakorn Pattayakorn	6610545766	Software and Knowledge Engineering	Engineering
Phasin Sae-Ngow	6610545383	Software and Knowledge Engineering	Engineering

Project Overview:

Main features

• Visualisations for historical weather data.
• View latest sensor readings.
• Predictions for future weather conditions.
• Event advisor to suggest if an event should be conducted and recommend equipment based on predicted weather conditions.

APIs provided

1. Weather statistics endpoints

Endpoint	Method	Description
/locations	GET	Return locations with observed weather data
/weather	GET	Return the Latest weather observation for the given location. If a date is specified return the closet observation to that date.
/weather/last	GET	Return weather observations from the last x days. If no duration is specified return weather observations from yesterday.
/weather/aggregate	GET	Return aggregated weather data from the last x days. If no duration is specified return weather data from yesterday.
/temperature/max	GET	Return the maximum observed temperature from the last x days. If no duration is specified return the maximum temperature from yesterday
/temperature/min	GET	Return the minimum observed temperature from the last x days. If no duration is specified return the minimum temperature from yesterday
/humidity/max	GET	Return the maximum observed humidity from the last x days. If no duration is specified return the maximum humidity from yesterday
/humidity/min	GET	Return the minimum observed humidity from the last x days. If no duration is specified return the minimum humidity from yesterday
/pressure/max	GET	Return the maximum observed pressure from the last x days. If no duration is specified return the maximum pressure from yesterday
/pressure/min	GET	Return the minimum observed pressure from the last x days. If no duration is specified return the minimum pressure from yesterday
/rainfall/max	GET	Return the maximum observed rainfall from the last x days. If no duration is specified return the maximum rainfall from yesterday
/rainfall/min	GET	Return the minimum observed rainfall from the last x days. If no duration is specified return the minimum rainfall from yesterday
/sensor/latest	GET	Return the Latest weather observation from the sensor.

2. prediction endpoints

Endpoint	Method	Description
/predict/temperature	GET	Predicts temperature values from the last observed date in the models training data until the specified date for the given location.
/predict/humidity	GET	Predicts humidity values from the last observed date in the models training data until the specified date for the given location.
/predict/pressure	GET	Predicts pressure values from the last observed date in the models training data until the specified date for the given location.
/predict/rain	GET	Predicts hourly rain from the last observed date in the models training data until the specified date for the given location. (Each returned value is either cloudy/no rain or rain)

3. Evaluation endpoints

Endpoint	Method	Description
/event/conditions	GET	Return the predicted weather conditions during the given period, maximum predicted temperature, highest heat index calculated, and rainy periods.
/event/describe	POST	Return a suggestion for hosting the event (yes/no), along with a description of important weather periods (heat wave, rainy periods) and suggested equipment.
/heatindex	GET/POST	When the method is GET, calculates a single heat index. Use POST for calculating multiple heat indices.

Installation

Please see Installation.md for installation instructions.

Using the application

Running the backend (API service)

1. create a virtual env.
   python -m venv env
2. Activate the virtual environment.
   

   # on Mac/Linux use
   .env/bin/activate
   
   # on Windows use
   .\env\scripts\activate
   

3. In the projects root directory, Run the backend using:
   python -m uvicorn api.app:app_api --port 8000 --reload
4. Use http://127.0.0.1:8000/explan/ as a base path to call the API
   or go to http://127.0.0.1:8000/docs/ to call the APIs using a GUI

Running the frontend (Web application + Visualizations)

1. Run the backend using the provided instructions.
2. navigate to the frontend directory.
3. Run the frontend using: streamlit run main.py
4. Wait for the site to load.
5. If the site does not open automatically, go to http://localhost:8501/ .

Database Schema

You will need a MySQL database to run this project.
you can find the required database schema in the database setup readme or in the project wiki.

Prediction model + data acquisition

The provided project already has starting models which are capable of weather prediction.
The provided data allows for predictions at Kasetsart University and the Nak Niwat 48 district. These models are (to a degree) capable of predicting weather conditions at other locations within Bangkok.

To allow the models to predict weather conditions at other locations, add the sample observation data in the api/models/trained_models/data directory.
(Note: The location data provided must be recorded in the Weather Data integration table)

Tutorials to create your own prediction models and sample code to collect data can be found here: khunakorn/ExtroPlanner-data