Optimal Japanese Restaurant Location In Manhattan

Introduction

• This exercise will seek to show a good location to open up a Japanese
  restaurant in Manhattan.
• This would be value to anyone looking to open up a Japanese
  restaurant based on certain criteria that is further discussed in the
  data section.

Data

• Number of any restaurant in each neighborhood
• Number of Japanese restaurants in each neighborhood
• Adjusted Gross Income (AGI)
• Zipcodes
• Latitude/Longitude (ie location of neighborhoods and restaurants)

Sources

• Foursqure API
• Source of file from 2014 NYU dataset that has coordinates of each New York City neighborhood
• IRS data to get AGI by zip: https://www.irs.gov/statistics/soi-tax-stats-individual-income-taxstatistics-2017-zip-code-data-soi
• geopy: this will enable mapping coordinates to zip code

Methodology & Exploration

• As mentioned earlier, we need to capture the following
  • Frequency of types of restaurants in each neighborhood
  • Number of any restaurant in each neighborhood
  • Number of Japanese restaurants in each neighborhood
  • Adjusted Gross Income (AGI).
    • Although we do can do this at various income levels, we are looking only at areas where
      the AGI is 75k+
• In addition, we will assess frequency of restaurants/top restaurants in
  each neighborhood
• Focus will be restricted to these neighborhoods within Manhattan

Using Foursquare API, We Can Pull In All Restaurants In Manhattan Within A Specified Range And Observe: There Are 2K+ Restaurants &122 Types Of Restaurants.

In Addition, We Can Assess Frequency By Top 10 Restaurants Within Each Neighborhood.

In The Dataset, We Can Determine Both The Number Of Japanese Restaurants And Total Restaurants.

We Can Pull In Adjusted Gross Income (AGI) By Zip Code. We Are Restricting This Further To AGI_Stub 3+, Which Indicates AGI >=75K.

Cluster Resulting Dataset Using K Means To Review Similar Neighborhoods.

…And Mapping The Results.

Results

One Cluster Has Little To None Japanese Restaurants Currently.

Other Clusters Have Many Japanese Restaurants And More Restaurants In General Too.

Discussion: Washington Heights, Chinatown, Clinton, Little Italy, and Soho Are The Better Areas To Consider. These Areas Are Clustered Together And Currently Have High Restaurant Counts And Low Japanese Restaurant Counts.

Conclusion

• Based On The Data We Used, There Is Opportunity To Open Up A Japanese Restaurant In Multiple Locations Within Manhattan
• Although Our Data Shows Potential Areas To Open A Japanese Restaurant, Additional Analysis Should Be Done To Address The Following
  • NYC Dataset Is From 2014
  • IRS Dataset Is From 2017
  • Is Foursquare Accurate (ie Are There More Japanese Restaurants In This Location)
  • Marketing: Does This Area Want Japanese Restaurants
• It Would Be Interesting To Add Additional Datapoints Such As Population Density, Race, etc