Exploratory Data Analysis 101: UFO Sightings Data

There has been a recent spike in UFO sightings around the US and the world. Yes, you read that right. And this has left some conspiracy theorists confused about who will be our next overlords - aliens or AI?

In the first couple of months of 2023, there seemed to be UFO-related news everywhere. The BBC reported that in 2022-23, US intelligence agencies received 500 reports of UFOs. Many of them were later dismissed as balloons and debris, but 500 is an astonishing number.

Now, whether you believe in aliens or not, UFO sightings make for some really interesting news. In this blog, let us try to find out if there are any patterns in UFO sightings around the world, and what is the best place and time to see a UFO.

About the UFO Sightings Data

This dataset contains over 80,000 reports of UFO sightings over the last century. You can access is here on Kaggle or view it on our data community (without logging in.)

We'll load and view this CSV file in Gigasheet, and start our data exploration. For reference, you can also integrate Gigasheet with databases, warehouses, and CRMs.

1) Most common UFO shapes

Is it a bird, is it a plane, or is it something else? Let us find out what the data has to say about UFO shapes. Grouping by 'shape', we get that there are 30 different UFO shapes, including unknown and other. Some fields are also blank.

Let us filter that out.

Now, we are left with 27 rows. Sorting them by row count, we see that 'light' is the most common UFO shape.

Let us plot this data into a chart. Select the fields you want and click on 'Chart data.'

2) Which is the best time to spot a UFO in the US?

Using similar grouping operations, we can see that the US has experienced the most number of UFO sightings (3192). So, before we figure out the date, we will filter the data to include rows where the location is set as the United states.  

Next, let us look at the column datetime which indicated the date the UFO was seen. This field is currently in the plain text data format. Before we start performing date operations, let us change the data type to date.

Converting it to date format:

‍

Whenever you convert a column to a different data type, Gigasheet adds another column labelled  <column name> as <new data type>. After this conversion, you will see another column, 'datetime as date.'

Next, we will use Gigasheet's explode date feature to break this date value into day, day of week, date, month, year, hour, and more.

‍

You will then be able to see various columns containing a single value for day, month, and year.

To find out which day of the year are we most likely to see UFOs, let us group by the day of the week. Right click on the column "datetime as date - Day of Week" and select 'Group.'

After we group the data, we can use 'Chart data' to display the data distribution with a bar graph.

So, as you can see, people are the most likely to see UFOs on a Saturday. If you don't have weekend plans yet, maybe you can consider a night of stargazing.

3) In Which Year Were The Most UFOs seen?

Let us now remove the location filter and then group by year. We will use the column 'datetime as date - Year.'

The year 2012 has the most number of UFO sightings. Makes sense, because 2012 was the year people believed that the world was going to end!

Do you ever wonder if aliens exist? We don't quite know yet. But you know what does exist? A no-code solution for processing large datasets no matter where they are stored. Say hello to Gigasheet.

Trust Gigasheet For Data Analysis

Gigasheet is a complete data analysis workbench designed to process large datasets with ease. Now, anyone can analyze large CSV, JSON, and XLSX files without spending hours learning to write and debug code.

Sign up for your FREE account here.