Sunday, December 25, 2016

Nairaland Christmas Birthday Analyzed and Visualized in Python

Members of nairaland.com forum that celebrates their birthday on Christmas dayHappy Birthday Christmas to this Nairaland forum members

Which member of Nairaland.com forum is celebrating his/her birthdays today? Let's derive some insights from the Birthday dataset.


Analysis of Nairaland members that share a birthday with Jesus Christ


Some months back as part of a "Data Science beginners tutorial",  I wrote a web crawler that extract the birthday details (username and age) of members of Nigeria largest online forum - Nairaland.com

Now talking about celebrating birthdays, there is something special about December, 25th I want to explore, to see at least if it is true with Nigerians...
According to a 2006 Harvard University study, Christmas birthdays are so rare. December 25 is least common day for births (after February 29).
I will use this post to verify above fact with Nigerians by using a public birthday list available on Nigeria largest online forum - Nairaland.com

We shall see if there are fewer Nigerians (nairaland members) celebrating there birthdays on this date! But first lets use our web crawler to extract and save the dataset in a csv file.

Extracting and saving the dataset in a csv file


I used the packages requests, BeautifulSoup and pandas. Below is the notebook containing the complete source code.



The Analysis


Let load the Christmas birthday list from csv into python and begin to explore it to reveal some insights in our journey to verify the above fact with respect to Nigeria.


Before I start any tech analysis, a simple glace at the summary statistics of the dataset show about 200 records. This is above the average number of birthdays for other days! I know this because in the past, I have collected the birthday list of other dates as seen below:

18/08/2016 = 113 birthdays

19/08/2016 = 98 birthdays
29/08/2016 = 82 birthdays
31/08/2016 = 87 birthdays

Another factor that was pointed out was Nigerians not been sincere with the birthdays especially when making casual registrations online. All these will hinder the progress of the analysis.

With this, we don't have to waste much time in drawing a conclusion the research.


Thanks for following.

1 comment: