README.md

Description

This project is the source code for Wildfire and Climate Change in US by Group 4 - Asia Pacific. With this project, it is possible to visualize historical wildfire data from 1992 to 2015. Moreover, some other factors that may affect wildfires are included, such as fuel, temperature, rain and wind. Finally, we included prediction results of our machine learning model.

Webapp screenshot:

System Requirements

To run this code example, you will need the following Python packages. It can be installed with pip install -r requirements.txt. The requirements.txt file can be found in the CODE folder.

Module	Version
pandas	1.1.3
flask	1.1.2
flask-socketio	4.3.1
numpy	1.19.1

Quick Start

Before running the code, make sure you download these datasets:

1. FPA_FOD_20170508.sqlite wildfire data in dataset/wildfire/FPA_FOD_20170508.sqlite. This file can be downloaded from Kaggle here.
2. The rest of the datasets folder can be downloaded here.

The datasets folder structure should be

datasets
   ├── fuel_moisture
   │   └── nfmd_compiled.csv
   ├── mlpreds
   │   └── ml_output.csv
   ├── temperature_and_precipitation
   │   └── tp_zipcode_county.csv
   ├── wildfire
   │   └── FPA_FOD_20170508.sqlite
   └── wind
       └── wind_with_fips.csv

To run the code, from this directory, type python3 main.py. Then, on a browser, go to 127.0.0.1:8080.

API

Weather forest data requires Professional Plan from weatherstack: https://weatherstack.com/

Run machine learning procedure

At app directory, run python3 api_download_ml.py

It will create a new ml_output.csv file in the datasets/mlpreds directory based on current datas' condition. This should be add as a daily procedure step on the application server.