lluis

CORRELACION

corrm <- cor(cat[,3:56])
catnum <- cat[,3:56]
#write.xlsx(corrm, file = "corr.xlsx")

HIER

hh <- dist(catnum, method = "euclidean")
hc <- hclust(hh, method = "ward.D2")
hc5 <- cutree(hc, k = 5)
table(hc5)

## hc5
##   1   2   3   4   5 
## 119  66 130 229  61

cat$hc5 <- hc5

#write_xlsx(cat, "catK.xlsx")

K1 <- subset(cat, hc5 == "1")
K1[,1:2]

## # A tibble: 119 x 2
##    CIUDAD  BARRIO                 
##    <chr>   <chr>                  
##  1 Austin  Barton Hills           
##  2 Austin  East Riverside         
##  3 Austin  Galindo                
##  4 Austin  Travis Heights         
##  5 Austin  Zilker                 
##  6 Bangkok Chinatown              
##  7 Bangkok Din Daeng/Huai Khwang  
##  8 Bangkok Monumento a la Victoria
##  9 Bangkok Phra Khanong           
## 10 Bangkok Thonburi               
## # ... with 109 more rows

K2 <- subset(cat, hc5 == "2")
K2[,1:2]

## # A tibble: 66 x 2
##    CIUDAD    BARRIO                  
##    <chr>     <chr>                   
##  1 Austin    Bouldin Creek           
##  2 Austin    South Lamar             
##  3 Bangkok   Bang Na                 
##  4 Bangkok   Chatuchak               
##  5 Bangkok   Klong Toey              
##  6 Bangkok   Ratchathewi/Phaya Thai  
##  7 Bangkok   Siam                    
##  8 Barcelona Nou Barris              
##  9 Barcelona Sant Pere/Santa Caterina
## 10 Berlín    Mitte                   
## # ... with 56 more rows

K3 <- subset(cat, hc5 == "3")
K3[,1:2]

## # A tibble: 130 x 2
##    CIUDAD    BARRIO         
##    <chr>     <chr>          
##  1 Austin    Centro         
##  2 Austin    Clarksville    
##  3 Austin    East Downtown  
##  4 Austin    North Loop     
##  5 Austin    Old West Austin
##  6 Bangkok   Aree           
##  7 Bangkok   Thong Lo       
##  8 Barcelona El Poble-sec   
##  9 Barcelona El Raval       
## 10 Boston    Brookline      
## # ... with 120 more rows

K4 <- subset(cat, hc5 == "4")
K4[,1:2]

## # A tibble: 229 x 2
##    CIUDAD  BARRIO           
##    <chr>   <chr>            
##  1 Austin  Dawson           
##  2 Austin  Hyde Park        
##  3 Austin  Parker Lane      
##  4 Austin  South Congress   
##  5 Austin  St. Edwards      
##  6 Austin  Upper Boggy Creek
##  7 Bangkok Banglampoo       
##  8 Bangkok Dusit            
##  9 Bangkok Pathum Wan       
## 10 Bangkok Sathorn          
## # ... with 219 more rows

K5 <- subset(cat, hc5 == "5")
K5[,1:2]

## # A tibble: 61 x 2
##    CIUDAD    BARRIO                         
##    <chr>     <chr>                          
##  1 Austin    Hancock                        
##  2 Austin    Universidad de Texas           
##  3 Bangkok   Upper Sukhumvit                
##  4 Barcelona El Born                        
##  5 Barcelona L'Antiga Esquerra de l'Eixample
##  6 Barcelona Les Corts                      
##  7 Barcelona Sant Andreu                    
##  8 Berlín    Prenzlauer Berg                
##  9 Boston    Allston-Brighton               
## 10 Boston    Cambridge                      
## # ... with 51 more rows

fviz_cluster(list(data = cat[,3:57], cluster = hc5))

plot(hc, cex = 0.6)
rect.hclust(hc, k = 5, border = 2:5)