Data Analysis on the Best Bars in New York City
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.
Image source: https://la.eater.com/2018/10/2/17928474/gold-line-bar-highland-park-stones-throw-records-photos-inside
Image source: | Motivation |
Many people go bars to connect, relax, have fun, and meet people. While others go to put an end to the monotonous life, stay in touch with friends, be seen, be heard, listen to music, watch games, etc. Whatever may be the intention of going, data shows bars provides social lubricant to relax people.
My sole intention of this project was to answer my friend’s question “Which is the best bar in New York City?” that I was unable to answer quantitatively when he asked me before his visit here. Prior to this project, I did not have any quantitative information regarding bars other than reviewing Yelp search results or other similar applications for bars. With this project I intended to update my understanding of bars around New York City with my own quantitative measurements .
| Questions expected to be answered |
What's the neighborhood in New York City has the most active night life? Which are the best bars in New York City? Which day of the week is best and worst to go bars? What percentage of bars are wheel chair accessible? What percentage of bars have happy hours, bar TV, own parking, etc.
| Methods and tools |
In order to collect data about bars in New York city, I scraped Yelp using Scrapy tool written in Python. All data cleaning, analysis and data visualization were performed in Pandas and NumPy. All of my coding including the data can be found in following git hub: link https://github.com/basantdhital/MY_pro
| Data on Neighborhoods with most active Night life |
Before diving into the best New York City bars, I wanted to find out which neighborhood had the most active night life in New York. To accomplish this, I created a bar plot demonstrating the number of bars versus neighborhood. From this bar plot, the top five neighborhoods with most active night life were found to be Mid town West, Mid town East, East Village, Upper East Side, and West Village.
| Data on Best Bars in New York City |
In order to find best bars in New York City, I created a "popularity index", defined by the product of the number of reviews and the bar ratings listed in the Yelp website. The best five bars in New York City on the basis of popularity index are shown below. Moreover, best bars were also found to have price range in the less expensive region.
| Best and worst night for going bars |
The best and worst night of a week to go bars were calculated from both the popularity index, and the best nights listed on individual bars page.
For example, if the bar has listed the best night to be Friday, it was given value 1 for Friday and rest of the days in week were given zero. Then values for the particular day of week was multiplied with the popularity index of each bar which was then summed over all the bars. Finally, whichever day of the week has the highest popularity index value was assigned the best night and that with the least value was worst night to go bars. The histogram of various night with popularity index is shown below:
| Other useful data |
From the data I collected from the Yelp website, I calculated various percentage of different facilities in bars. In New York city, 30 percent bars are wheel chair accessible. Only 8.15 percent have bar dancing facility and 9.44 percent have their own parking garage. The percentage of bars that provide reservations, happy hours, and with bar TV are 51.23, 53.04, and 53.56, respectively.
| Conclusions |
I hope with these informations about bars will be helpful to choose your best bars in New York City. From business point of view, this project provides areas to improve such as bar parking, bar dancing, etc. in order to have successful bars in New York City.
| Future directions |
It would be nice to additionally collect more information about male to female ratio in each bar by scraping the individual reviewers for each bar. Male to female ratio might help people to choose right bar to go according to their interest. Finding zones of popular drinking site (may be using heat map) might provide driving industry new area to focus to expand their business in future.
About Author
Basant Dhital
View all posts by Basant Dhital >Leave a Comment
