In this repository a random forest regression model was used to find the probability of a fire occuring in a certain region based on weather conditions. The purpose of each file is as follows:

2015_to_2020_Daily.xlsx - Provides daily weather conditions (such as maximum temperature and average wind speed) for over 150 weather stations across the state of California (from NOAA)

3DS_notebook.ipynb - uses the information contained in the weather_data_with_fires.csv to generate a random forest regression model and a variety of graphs such as the importance score of each feature

CA_State.shp - Provides California state boundary lines (from CA.gov)

California_Fire_Perimeters_2015_to_2020.shp - Provides historical fire data in the state of California

Fire_Area.ipynb - Maps the historical fire data on the state of California

Voronoi_Area_v3.ipynb - Maps the area represented by each weather station using a voronoi diagram

Weather_with_Fires.ipynb - Combines the fire and voronoi maps together to generate a .csv file containing the weather data and any fire occurences (denoted by both name and a 1 if a fire occurred, otherwise 0)

weather_data_with_fires.csv - Provides the .csv file generated by the previous jupyter notebook which was then used in 3DS_notebook.ipynb to create the regression model