Research Objective

Identify factors which the hospital should take into account while placing the billboards across the city.

Rationale behind choosing the city

Delhi , being capital, is one of the most densely populated cities in the country. Air pollution index suggests that the inhabitants of the city is facing serious health issues. Majority of the population is facing bad cardiac health issues. All this makes advertising in the city good for the business as:-

Due to large population, the conversion rate is comparatively higher than any other cities in the country.
Delhi has 153 micrograms of PM2.5 per cubic metre which is causing pollution at alarming level.

Proxies used

Subway station
Bus Station
Liquor store
Cemetery

Reasons for choosing the proxies

Subway/metro station - A large chunk of daily commuters involve office goers , offices located in and around places(cyber city, noida, Manesar) which are very densely polluted due to large amount of vehicles coming in daily. Therefore making these office goers prone to health disease due to physical inactivity which further amounts to stress , and other health issues . There are good chances of these set of people in availing these services.
Bus station - Wider audience who look for affordable health solutions. Making them aware of these services could benefit in a way , as many of them may not be aware of these services.
Funeral home/cementry - vulnerable state of mind. Putting hoardings around these locations will help generate awareness in people about the things which can be done in life threatening, devastating times.
Liquor store - Placing billboards at these locations will hit the psychological mind set of the buyer regarding health awareness, as alcohol consumption lead to health problems.

Code

library("RCurl")
library("jsonlite")
library("plotGoogleMaps")
library("geosphere")
 
key = "" # Please input the key here.

# malls search
#url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=malls+in+delhi&types=shopping_mall&location=28.6139,77.2090&radius=25000&key=",key)
#doc <- getURL(url)
#x <- jsonlite::fromJSON(doc)
#malls = x$results$geometry$location
#head(malls)

# hospitals
#url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=hospitals+in+delhi&types=hospital&location=28.6139,77.2090&radius=25000&key=",key)
#doc <- getURL(url)
#x <- jsonlite::fromJSON(doc)
#hospitals = x$results$geometry$location
#head(hospitals)

# banks
#url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=banks+in+bangalore&types=bank&location=28.6139,77.2090&radius=50000&key=",key)
#doc <- getURL(url)
#x <- jsonlite::fromJSON(doc)
#banks = x$results$geometry$location
#head(banks)


# health
#url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=health+in+delhi&types=health&location=28.6139,77.2090&radius=25000&key=",key)
#doc <- getURL(url)
#x <- jsonlite::fromJSON(doc)
#health = x$results$geometry$location
#head(health)

# Bus station search
url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=bus_station+in+delhi&types=bus_station&location=28.6139,77.2090&radius=25000&key=",key)
doc <- getURL(url)
x <- jsonlite::fromJSON(doc)
busstation = x$results$geometry$location
head(busstation)

#url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=train_station+in+delhi&types=train_station&location=28.6139,77.2090&radius=25000&key=",key)
#doc <- getURL(url)
#x <- jsonlite::fromJSON(doc)
#trainstation = x$results$geometry$location
#head(trainstation)

# Subway station search
url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=subway_station+in+delhi&types=subway_station&location=28.6139,77.2090&radius=25000&key=",key)
doc <- getURL(url)
x <- jsonlite::fromJSON(doc)
subwaystation = x$results$geometry$location
head(subwaystation)

# Cemetery search
url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&query=cemetery+in+delhi&types=cemetery&location=28.6139,77.2090&radius=25000&key=",key)
doc <- getURL(url)
x <- jsonlite::fromJSON(doc)
cemetery = x$results$geometry$location
head(cemetery)

# Liquor store  search
url = paste0("https://maps.googleapis.com/maps/api/place/radarsearch/json?&type=liquor_store&location=28.6139,77.2090&radius=25000&key=",key)
doc <- getURL(url)
x <- jsonlite::fromJSON(doc)
liquorstore = x$results$geometry$location
head(liquorstore)




#malls$type = "Mall"
#banks$type = "Bank"
#health$type = "health"
#hospitals$type = "Hospital"
busstation$type="bus_station"
#trainstation$type = "train_station"
subwaystation$type= "subway_station"
cemetery$type= "cemetery"
liquorstore$type="liquor_store"
data = rbind(busstation,subwaystation,cemetery,liquorstore)
dim(data)

write.csv(data,"delhi_places.csv", row.names = F)

###################################################################
###################################################################
###################################################################

# let's plot the bus stations and do clustering based on distance matrix

sample = busstation
coordinates(sample) <-~ lng +lat # Create cordinates
proj4string(sample) = CRS('+proj=longlat +datum=WGS84') # Add Projections

m<-mcGoogleMaps(sample,zcol = "type", mapTypeId='ROADMAP') # Plot on Google maps


# let's plot the subway stations and do clustering based on distance matrix
sample = subwaystation
coordinates(sample) <-~ lng +lat # Create cordinates
proj4string(sample) = CRS('+proj=longlat +datum=WGS84') # Add Projections

m<-mcGoogleMaps(sample,zcol = "type", mapTypeId='ROADMAP') # Plot on Google maps

# let's plot the cemetery and do clustering based on distance matrix
sample = cemetery
coordinates(sample) <-~ lng +lat # Create cordinates
proj4string(sample) = CRS('+proj=longlat +datum=WGS84') # Add Projections

m<-mcGoogleMaps(sample,zcol = "type", mapTypeId='ROADMAP') # Plot on Google maps

# let's plot the liquor store and do clustering based on distance matrix
sample = liquorstore
coordinates(sample) <-~ lng +lat # Create cordinates
proj4string(sample) = CRS('+proj=longlat +datum=WGS84') # Add Projections

m<-mcGoogleMaps(sample,zcol = "type", mapTypeId='ROADMAP') # Plot on Google maps





# Get the coordinates
p2 = busstation[,1:2]

# calculate distances
dist_mat = matrix(0,nrow(p2),nrow(p2))

for (i in 1:nrow(p2)){
  for (j in 1:nrow(p2)){
    dist_mat[i,j] = distCosine(p2[i,],p2[j,], r=6378173)/1000    
  }
}

class(dist_mat)
dist_mat[1:10,1:10]

# Create clusters based in distances
fit <- hclust(as.dist(dist_mat), method="ward")
plot(fit) # display dendogram

groups <- cutree(fit, k=18) # cut tree into 18 clusters
# draw dendogram with red borders around the 18 clusters
rect.hclust(fit, k=18, border="red") 

sample$group = groups # Assign cluster groups

# Plot stores with clustor as label
m <- mcGoogleMaps(sample, mapTypeId='ROADMAP', zcol="group")


#########################################################
#########################################################
#########################################################

sample = data
coordinates(sample) <-~ lng +lat # Create cordinates
proj4string(sample) = CRS('+proj=longlat +datum=WGS84') # Add Projections

m<-mcGoogleMaps(sample,zcol = "type", mapTypeId='ROADMAP') # Plot on Google maps

Results of analysis

Note: The billboard locations are highlighted in red circles.

Hauz khas - Outer ring road / Aurobindo Marg:

All the four targeted proxies lie in this area, making the area highly likely to be visible to our target segments. Regular movement of the traffic ensures that the ad is visible to a large chunk of population. Hauz khas

Cannaught Place - Inner circle

The cluster in CP comprises of multiple metro stations and liquor store. This contributes to our target population, and therefore placement of bill boards at these locations will have a huge impact.

Rajouri garden - Raja Garden Chowk

It is one of the most densely populated area of Delhi, with local population in majority. Bill boards in area will help target a) Local Delhites prone to bad health because of the high pollution levels in this area. 2) People taking bus and metro for commuting, and therefore making them aware of these services. $Raja Garden Chowk$

Intersection of Noida Link road and NH-24

Metro station, bus stations, Cementry lies equidistant from this location. These places brings our target population’d view on the bill boards placed a this area.

Group Name - Team Analytics

71610025 - BASAVA ESWARA KRISHNA ALURI

71610076 - SHANTNU GUPTA

71610032 - KARTIK MAHESHWARI

71610038 - MOHIT KALSI

City of choice

DELHI

Client

City Hospital

Decision problem