Chun Shui Tang Map Clustering

To install packages, enable KPI, and prepare data

install.packages('googleway')

install.packages("writexl")

install.packages('geosphere')

library(googleway)

library(writexl)

library(dplyr)

library(geosphere)

## Enabling API on Google Cloud

myApi = 'AIzaSyCzMc0FVwsEbSGEsBvTzQKY59UsnEN9fNE'

## Finding All of Chun Shui Tang Branches

search_str <- google_places(search_string = 'Chun Shui Tang in Taiwan', 

       location=c(23.899504, 120.981795), 

       radius=50000, key=myKpi)

round2 <- google_places(search_string = 'Chun Shui Tang in Taiwan', 

            location=c(23.899504, 120.981795), 

            radius=50000, key=myKpi,

            page_token = search_str$next_page_token)

round3 <- google_places(search_string = 'Chun Shui Tang in Taiwan', 

            location=c(23.899504, 120.981795), 

            radius=50000, key=myKpi,

            page_token = round2$next_page_token)

## Keeping Results

result <- search_str$results

result2 <-round2$results

result3 <-round3$results

## Clustering Part

#Mean of Lat Lon

MeanLat <- mean(final_df$lat, na.rm = TRUE)

MeanLon <- mean(final_df$lng, na.rm = TRUE)

#Hierarchical Clustering

hc <- hclust(DMat, method="complete")

final_df$Clusters <- cutree(hc, k = 8)

Convert data to the appropriate data frame

## Binding Back Latitude and Longitude Data

result <- search_str$results %>%

 cbind(., search_str$results$geometry$location)

result2 <- round2$results %>%

 cbind(., round2$results$geometry$location)

result3 <- round3$results %>%

 cbind(., round3$results$geometry$location)

## Convert to Dataframe

df <- as.data.frame(result)

df2 <- as.data.frame(result2)

df3 <- as.data.frame(result3)

Export to Excel file

# Specify the corrected file path with double backslashes or forward slashes

file_path <- "C:/Users/Phetcharee/Desktop/CSTProject/chun_shui_tang_branches.xlsx"

# Write final_df to Excel file

write_xlsx(final_df, path = file_path)

To generate the distance matrix table

#Distance of all stores from mean lat lon

CSTLatLon <- final_df[,c(17,18)]

#Distance Matrix 

Distance_Mat<- distm(CSTLatLon[2:1],CSTLatLon[2:1],

           fun = distHaversine)

Distance_Mat<- as.data.frame(Distance_Mat)

Distance_Mat[is.na(Distance_Mat)]<-0

DMat<- as.dist(Distance_Mat)

# Bind data frames together using dplyr's bind_rows function

final_df <- bind_rows(df, df2, df3)

Based on my observation of the location points on the map, Chun Shui Tang has three main areas where its stores are located. The first area is Taichung, which has ten stores with an average customer rating of 4.06. Additionally, the Chun Shui Tang Siwei original store has the most reviews on Google Maps.

The brand has eight stores in the Taipei area, with an average rating score of 3.90. After reviewing customer complaints, it appears that the main issues are the waiting times at each branch and the quality of the services provided, which the brand needs to improve.

In the Taoyuan area, Chun Shui Tang has only two operating stores with an average rating score of 3.95. However, these two stores have the lowest user rating numbers, getting only 3.34% of all customers who review on Google Maps.