rachelewooten/stormeventsdataproject

Examining the NOAA Storm Events Database for the human costs of weather.

StormEventsDataProject

Examining the NOAA Storm Events Database for the human costs of weather. The current model examines the storm event details from 1996-2018, as prior data did not include most types of storm event. Future analysis will try to incorporate location and specific fatality information to improve models. In addition, population census data would be valuable for evaluating relative risk.

Data pre-cleaning: In order to examine the financial costs of individual storm events, the DAMAGE_PROPERTY and DAMAGE_CROPS columns need to be cleaned up. They are currently input in the form "100.0", "2.5K", ".2B". The function functions.convert_num( x) converts a string of any of these forms into a float value.

Note: None of these evaluations incorporate inflation yet.

---

1. Scrape

First, run python scrape.py in $(PROJECT_HOME) folder to scrape the .csv files from the NOAA Storm Events Database into project home folder. The .csv files will be stored in 3 folders: $(PROJECT_HOME)/details, $(PROJECT_HOME)fatalities, $(PROJECT_HOME)/locations

2. Clean

In $(PROJECT_HOME), run

bash initialize_db.sh
python clean.py postgresql_username postgresql_password

in order to create database and upload storm-events details after 1996