Do you have a Dog? — How much do you know about them?
This is a data story about various dog breeds.
Dogs are the only mammal on earth having the most variable number of breeds. The Fédération Cynologique Internationale (FCI), list about 360 breeds globally, the Labrador retriever being the most popular of them all based on a rating by the American Kennel Club in 2021.
In this article, I explored a dataset on dog breeds to uncover insights about some of the:
- Top 10 Most common dog breeds
- Top 10 most loved dog breeds
- Top 10 dog breeds by rating
The Data
The analysis was based solely on a dataset gotten from a twitter account (@WeRateDogs) that rate dogs based on the Dog’s pictures posted. This data was then augmented the result of a machine learning algorithm that predicts dogs’ breeds based on the its picture. You can learn more about the neural network algorithm here.
You can read this article if you are interested in seeing how the data wrangled or you can choose to download the final dataset.
1. Top 10 Most common dog breeds
According to the American Kennel Club (AKC), the Labrador Retriever has held the top spot since 1991. So let’s compare our findings to the AKC’s 2021 Top 5.
- Labrador Retriever
- French Bulldog
- Golden Retriever
- German Shepherd Dog
- Poodle
Based on @WeRateDogs data, here are the top 10 most common breed.
What do you think? I think it is fair enough. At least our top 2 made the list. This insight is even clearer in the below word-cloud visualization of the dog breeds.
2. Top 10 most loved dog breeds
The favorite_count is the number of likes each dog posted got for that tweet. Since we are looking at the breeds as a whole and not the individual dog likes. I grouped the data by their breeds to find the average number of likes per breed.
So here is the result:
The multitude does not matter after all, only 1 Great Pyrenees is enough to win the heart of @weratedog’s followers. Anyways, let us see what our most loved dog looks like.
By the way, we could also use the retweet_count to determine the most loved dog but the below visualization shows that our result is not likely to be different. Since there is a strong positive correlation between the favorite_count and the retweet_count.
3. Top 10 dog breeds by rating
Disclaimer : @WeRateDogs has the exclusive right to determine the ratings. So in case your dog’s breed doesn’t make this list, you can banter with them on twitter @WeRateDogs.
Takeaways
The 3 major takeaways from this article are:
- The most common dog breed is a retriever, whether Golden or Labrador.
- The most loved dog breed by @WeRateDogs’ twitter followers is a little Great Pyrenees
- The most rated Dog breed is a Border collie that loves ice cream.
Limitations: here are some of the limitation this analysis:
- The ratings are biased and subject to the discretion or mood of the @WeRateDogs account administrator.
- The number of dogs per breeds in the dataset are not evenly distributed. Hence, using the average of the retweets, likes (favorite_count) & ratings tends to favour the breeds with fewer number of dogs. However, we have compensated them in the Dogs gallery section.
- The analysis only explored only basic insights from summary statistics, further relationships can be explored or draw inferences from the analysis.
Nonetheless, this is an interesting start to python for data analysis. Many thanks to ALX africa & Udacity.
Thanks for reading.