Data

Origin of the Data (Where did it come from?):

We pulled all the information we needed about the hurricanes in the Atlantic and Pacific Ocean respectively from a database maintained by the National Hurricane Center. Here is a link to this data: https://www.nhc.noaa.gov/data/#hurdat

Inside this database are hurricanes from the Atlantic Ocean from 1851 to 2018 while in the Pacific our range of dates went from the year 1949 to 2018.

Along with this link we were given a six-page PDF on how to read the data and what each number in the file meant in terms of either the hurricane type, wind speed, pressure, and location. Here is a link to this document: http://www.aoml.noaa.gov/hrd/hurdat/hurdat2-format-may2015.pdf

Obviously we didn't just use the original format in which the data was given to us. We created an easier way to view the tables for hurricanes in each ocean. So we created a bunch of names for each column, which specified exactly what every number or letter meant in the data table. Below is a screenshot of each of the column names and where they belong in the data set

Mutating and adding new columns:

After creating the columns 1 and 2 (date and time), we decided it would be much easier to make that one general column to lessen the confusion when looking at the data tables. So we combined the Date and Time columns to make a new "Timestamp" column, which had a data type of ddtm, which stands for date-time. Another reason for doing this was to make it easier to create visualizations using this variable type and to make the data look more clean.

As we progressed further into the project we realized it would also be easier to create the categorization graph by adding a column in the data table which specified under which category a certain hurricane fell under. Using this guide found in the NHC's website (https://www.nhc.noaa.gov/aboutsshws.php), we were able to find a way to categorize each hurricane. We created a function which looped through every hurricane's maximum wind speed and depending on what the number was, it fell into a certain category. Then we created a column called "Category" and labeled each hurricane from 1 through 5 depending on the wind speed. However we did run into some hurricanes which had max speeds which did not fall into the range specified by the National Hurricane Center so we labeled them as Category 0 Hurricanes.

In addition to these two changes to the original data table, we also created a function which assigned a generic name to a hurricane if it didn't have one. The name created depended on where the hurricane was found (Atlantic or Pacific) along with a unique ID.

We also changed the latitude and longitude coordinates to make them friendlier to map on Leaflet, by making our own function called "remake_coordinates". In this function we changed the coordinates to the four basic directions of North, East, South, and West.

Below is what our data tables looked like. There is one example for Pacific and one for Atlantic Hurricanes each. Since we had numerous columns, multiple screenshots were needed here.