- Jacob McKenney
- Luke Sala
This project uses data found on https://www.spc.noaa.gov/wcm/#data about tornadoes from 1950-2020 to do analysis and answer questions and predict tornado trends.
-analysis-results.txt: Contains output from running analysis functions
dataprocessing.py: Imports, retrieves, cleans, filters, and joins datasetsanalysis.py: Runs analysis on dataframes retrieved fromdataprocessing.pytuning_args.py: Stores static tuning parameters for different tuning variationsdata: Folder containing all data that is imported or written during analysis. Also includes a pdf explaining the tornado dataset usedfigures: Folder containing figures that are produced by runninganalysis.pytesting.py: Script that tests our functions against a smaller datasettest_data: Directory holding test filestest_figures: Figures created by our testing scriptfinal_report.pdf: Our report for our analysis
Make sure to have these packages installed in the virtual environment you
are running analysis.py
- pandas
- geopandas
- seaborn
- matplotlib.pyplot
- sys
- os
- sklearn
- Census (install with
conda install -c conda-forge census) - shapely.geometry
Our script optionally takes in an argument via the command line which
controls whether or not to use the full tuning parameters for tuning the
machine learning models. If you run the script with no arguments it
will use a less comprehensive set of tuning parameters to improve runtime. To
run the full tuning parameters use the following command:
python analysis.py True
Where python is the path to the proper python executable for your env.
Else just run:
python analysis.py
An example run command for our env is: /Users/jacobmckenney/opt/anaconda3/envs/cse163/bin/python /Users/jacobmckenney/Desktop/cs/cse163/project/tornado_analysis/src/analysis.py True
IMPORTANT: make sure you are in the src directory under tornado-analysis when running the script