Data Exploration

Dataset Snapshot

• Below is a snapshot of the hurricane dataset (source from NOAA, and dataset obtained from GitHub)

• The dataset covers data through 2015.

Data Processing and Aggregation

In this dataset, I have hurricane data by ID (indicating each hurricane), by date and time (4 data points per day, every six hours) and location level.  In my other article, I demonstrated the geo level visualization. This article will only focus on the data visualization without geospatial information. I'd like to show you the data aggregation to yearly level and chart the information I've extracted from them.

Python package 'pandas' provides a very powerful way to preprocess the dataset and aggregate it to my desired level. Here I created two data summary as below:

1. Top Wind speed (mph) by year. 

  2. Number of hurricanes formed by year.

Data Visualization using Matplotlib

There are many different ways to plot your data. You can plot one chart at a time. Below is an example of creating one chart with the size of 15 by 5, and creating a bar plot to show the total number of hurricanes formed each year. 

Of course, you are able to change the color, edit the name of the axis, etc.

And many times, you may want to combine information in one chart, or create a few subplots within one big canvas. The below plot is an example where I created two charts in one plot, side by side. The top plot is where I combined the hurricane counts and the top wind speed information per year in one chart. And the bottom plot is the bar chart showing the counts by year.