Unsupervised Machine Learning

PCA - Iris

Species - 범주

나머지 - 연속형

head(iris)

##   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
## 1          5.1         3.5          1.4         0.2  setosa
## 2          4.9         3.0          1.4         0.2  setosa
## 3          4.7         3.2          1.3         0.2  setosa
## 4          4.6         3.1          1.5         0.2  setosa
## 5          5.0         3.6          1.4         0.2  setosa
## 6          5.4         3.9          1.7         0.4  setosa

2. 결측치 확인

## Sepal.Length  Sepal.Width Petal.Length  Petal.Width      Species 
##            0            0            0            0            0

3. 변수별 기술 통계 및 분포 확인

중위수와 평균의 차이가 크면, 이상치가 많다는 의미로 볼 수 있다. (중요)

summary(iris)

##   Sepal.Length    Sepal.Width     Petal.Length    Petal.Width   
##  Min.   :4.300   Min.   :2.000   Min.   :1.000   Min.   :0.100  
##  1st Qu.:5.100   1st Qu.:2.800   1st Qu.:1.600   1st Qu.:0.300  
##  Median :5.800   Median :3.000   Median :4.350   Median :1.300  
##  Mean   :5.843   Mean   :3.057   Mean   :3.758   Mean   :1.199  
##  3rd Qu.:6.400   3rd Qu.:3.300   3rd Qu.:5.100   3rd Qu.:1.800  
##  Max.   :7.900   Max.   :4.400   Max.   :6.900   Max.   :2.500  
##        Species  
##  setosa    :50  
##  versicolor:50  
##  virginica :50  
##                 
##                 
## 

boxplot(iris[,1:4])

prcomp(iris[,1:4], 
       center = T, 
       scale. = T) ->iris.pca  #scale 평균 0, 분산 1 일때 처리해주기 

summary(iris.pca)

## Importance of components:
##                           PC1    PC2     PC3     PC4
## Standard deviation     1.7084 0.9560 0.38309 0.14393
## Proportion of Variance 0.7296 0.2285 0.03669 0.00518
## Cumulative Proportion  0.7296 0.9581 0.99482 1.00000

   Standard Deviation : 제곱 = 분산 = eigenvalue 
   Proportaion of Variance : 전체 분산에서 차지하는 비율 
   Cumlative Proportion : 누적 비율 
   
   
   PC 1은 72.9% 전체 변동량을 설명한다 
   PC 1 + PC 2 는 95.81%를 설명한다. 

Rotation

각 주성분들의 eigenvector (= 각 변수들의 가중치)

iris.pca$rotation

##                     PC1         PC2        PC3        PC4
## Sepal.Length  0.5210659 -0.37741762  0.7195664  0.2612863
## Sepal.Width  -0.2693474 -0.92329566 -0.2443818 -0.1235096
## Petal.Length  0.5804131 -0.02449161 -0.1421264 -0.8014492
## Petal.Width   0.5648565 -0.06694199 -0.6342727  0.5235971

X

head(iris.pca$x)

##            PC1        PC2         PC3          PC4
## [1,] -2.257141 -0.4784238  0.12727962  0.024087508
## [2,] -2.074013  0.6718827  0.23382552  0.102662845
## [3,] -2.356335  0.3407664 -0.04405390  0.028282305
## [4,] -2.291707  0.5953999 -0.09098530 -0.065735340
## [5,] -2.381863 -0.6446757 -0.01568565 -0.035802870
## [6,] -2.068701 -1.4842053 -0.02687825  0.006586116

Scree Plot

plot(iris.pca, 
     type='l', 
     main = 'scree plot')

1. 시각화 기준 2 or 3 개
2. Eigenvalue > 1 기준으로
3. Scree plot 에서의 elbow point 를 설정한다.

3 번쨰 주성분까지 선택하면 안되는 이유 -> Variance 는 eigenvalue 인데 1 이하 임으로, 2개의 주성분이 적당하다.

4. 차원 축소

2 개의 차원으로 축소

head(iris.pca$x[,1:2], 10)

##             PC1         PC2
##  [1,] -2.257141 -0.47842383
##  [2,] -2.074013  0.67188269
##  [3,] -2.356335  0.34076642
##  [4,] -2.291707  0.59539986
##  [5,] -2.381863 -0.64467566
##  [6,] -2.068701 -1.48420530
##  [7,] -2.435868 -0.04748512
##  [8,] -2.225392 -0.22240300
##  [9,] -2.326845  1.11160370
## [10,] -2.177035  0.46744757

5. 2 차원 시각화

autoplot(iris.pca, 
         data=iris,
         colour= 'Species')

6. 고급 시각화 -1

fviz_pca_biplot(iris.pca, 
                # Individuals
                geom.ind = "point",
                fill.ind = iris$Species, col.ind = "black",
                pointshape = 21, pointsize = 2,
                palette = "jco",
                addEllipses = TRUE,
                # Variables
                alpha.var ="contrib", col.var = "contrib",
                gradient.cols = "RdYlBu",
                
                legend.title = list(fill = "Species", color = "Contrib",
                                    alpha = "Contrib"))

고급 시각화-2

fviz_pca_biplot(iris.pca, 
                # Fill individuals by groups
                geom.ind = "point",
                pointshape = 21,
                pointsize = 2.5,
                fill.ind = iris$Species,
                col.ind = "black",
                # Color variable by groups
                col.var = factor(c("sepal", "sepal", "petal", "petal")),
                
                legend.title = list(fill = "Species", color = "Clusters"),
                repel = TRUE) +
  ggpubr::fill_palette("jco")

V 가까운 거리/방향성 = 상관성 증가 PC1 (DIM1) 기준 좌우 / PC2(DIM2) 기준 상하

PCA - AirBnB

2008년 이후, 게스트와 호스트는 여행이나, 좀 더 색다르고, 개별적인 여행 경험을 얻고자 에어비엔비를 사용하였다.

Column Books

idlisting: ID name: name of the listing host_id: host ID host_name: name of the host neighbourhood_grouplocation: neighbourhoodarea latitude: latitude coordinates longitude: longitude coordinates room_typelisting: space type price: price in dollars minimum_night: samount of nights minimum number_of_reviews: number of reviews last_review: latest review reviews_per_month: number of reviews per month calculated_host_listings_count: amount of listing per host availability_365: number of days when listing is available for booking

Data Pre Processing

setwd("C:/Users/Administrator/Desktop/R Analysis/Business R Aanlysis source")

read.csv("AB_NYC_2019.csv") -> df

str(df)

## 'data.frame':    48895 obs. of  16 variables:
##  $ id                            : int  2539 2595 3647 3831 5022 5099 5121 5178 5203 5238 ...
##  $ name                          : chr  "Clean & quiet apt home by the park" "Skylit Midtown Castle" "THE VILLAGE OF HARLEM....NEW YORK !" "Cozy Entire Floor of Brownstone" ...
##  $ host_id                       : int  2787 2845 4632 4869 7192 7322 7356 8967 7490 7549 ...
##  $ host_name                     : chr  "John" "Jennifer" "Elisabeth" "LisaRoxanne" ...
##  $ neighbourhood_group           : chr  "Brooklyn" "Manhattan" "Manhattan" "Brooklyn" ...
##  $ neighbourhood                 : chr  "Kensington" "Midtown" "Harlem" "Clinton Hill" ...
##  $ latitude                      : num  40.6 40.8 40.8 40.7 40.8 ...
##  $ longitude                     : num  -74 -74 -73.9 -74 -73.9 ...
##  $ room_type                     : chr  "Private room" "Entire home/apt" "Private room" "Entire home/apt" ...
##  $ price                         : int  149 225 150 89 80 200 60 79 79 150 ...
##  $ minimum_nights                : int  1 1 3 1 10 3 45 2 2 1 ...
##  $ number_of_reviews             : int  9 45 0 270 9 74 49 430 118 160 ...
##  $ last_review                   : chr  "2018-10-19" "2019-05-21" "" "2019-07-05" ...
##  $ reviews_per_month             : num  0.21 0.38 NA 4.64 0.1 0.59 0.4 3.47 0.99 1.33 ...
##  $ calculated_host_listings_count: int  6 2 1 1 1 1 1 1 1 4 ...
##  $ availability_365              : int  365 355 365 194 0 129 0 220 0 188 ...

df %>%  
  select(-reviews_per_month, -last_review, -latitude, -longitude) %>%  
  mutate(name = as.character(name), 
         id = as.character(id), 
         host_id= as.character(host_id), 
         hots_name = as.character(host_name),
         price = as.numeric(price)) -> df_2

Numeric 만 추출

df_num <-df_2 %>% 
  select_if(is.numeric)



df_non_numeric <-df_2 %>% 
  select_if(Negate(is.numeric))

Descriptive Statistics

round(stat.desc(df_num, basic=F), digit=3) ->statistic_table


datatable(statistic_table)

boxplot(df_num)

Correlation

cor(df_num, method="pearson")

##                                      price minimum_nights number_of_reviews
## price                           1.00000000     0.04279933       -0.04795423
## minimum_nights                  0.04279933     1.00000000       -0.08011607
## number_of_reviews              -0.04795423    -0.08011607        1.00000000
## calculated_host_listings_count  0.05747169     0.12795963       -0.07237606
## availability_365                0.08182883     0.14430306        0.17202758
##                                calculated_host_listings_count availability_365
## price                                              0.05747169       0.08182883
## minimum_nights                                     0.12795963       0.14430306
## number_of_reviews                                 -0.07237606       0.17202758
## calculated_host_listings_count                     1.00000000       0.22570137
## availability_365                                   0.22570137       1.00000000

PCA Analysis

prcomp(df_num, 
       center=T,
       scale=T) ->pca_bab


summary(pca_bab)

## Importance of components:
##                           PC1    PC2    PC3    PC4    PC5
## Standard deviation     1.1695 1.0646 0.9845 0.9334 0.8115
## Proportion of Variance 0.2735 0.2267 0.1938 0.1742 0.1317
## Cumulative Proportion  0.2735 0.5002 0.6940 0.8683 1.0000

PC 4가 86.83%의 변동률을 설명하는 것으로 4개가 적당해보인다.

plot(pca_bab, 
     type='l', 
     main = 'scree plot')

X

head(pca_bab$x[,1:5], 5)

##             PC1        PC2         PC3         PC4        PC5
## [1,]  1.0032784  0.5813873  0.07919756  0.30552371 -1.5550835
## [2,]  1.0039239  1.1772506  0.45957024  0.03989221 -1.1041550
## [3,]  0.9551646  0.4063914  0.06127306  0.17283789 -1.7050652
## [4,]  0.2656214  5.0369605  0.35645301 -0.85687593  2.2226285
## [5,] -0.6608942 -0.5561055 -0.32651731 -0.15276554  0.3290961

Rotation - eigenvectors

pca_bab$rotation

##                                       PC1        PC2          PC3         PC4
## price                          0.27569841 -0.2114304  0.920046259 -0.12497146
## minimum_nights                 0.46486745 -0.2822177 -0.335773478 -0.73645904
## number_of_reviews              0.03681049  0.8354198  0.083556017 -0.18126643
## calculated_host_listings_count 0.57439556 -0.1461332 -0.183649326  0.63909258
## availability_365               0.61368019  0.3954354  0.007897413  0.02670907
##                                       PC5
## price                           0.1310900
## minimum_nights                  0.2216658
## number_of_reviews               0.5107637
## calculated_host_listings_count  0.4544760
## availability_365               -0.6828263

    해석) PC1 365일 중, 가용 가능 날짜가 positive correlation 
          PC2 리뷰의 숫자가 가장 positve correlation 하다. 
          PC3 가격
          PC4 최소 숙박일이 낮아지면, listing_count 은 증가한다. 

sdev

pca_bab$sdev

## [1] 1.1694633 1.0645600 0.9845079 0.9334173 0.8115071

plot(summary(pca_bab)$importance[3,])

fviz_eig(pca_bab)

fviz_pca_var(pca_bab, col.var = "black")

숫자형 변수만 추출하여, PCA 분석 결과, 전체 변동 중 50%의 변동은 PC1 (27.35%) 와 PC2 (22.67%)를 차지하고 있다. PC1 에서 listing_count 와 availability_365가 가장 많은 변동량을 설명하고 있고, PC2에서는 review의 갯수이다. 그 외 다른 숫자형 변수 들은 약간의 다중 공선성이 서로간 있어보인다.

fviz_pca_ind(pca_bab, col.ind="cos2", geom = "point", gradient.cols = c("white", "#2E9FDF", "#FC4E07" ))

summary(pca_bab)

## Importance of components:
##                           PC1    PC2    PC3    PC4    PC5
## Standard deviation     1.1695 1.0646 0.9845 0.9334 0.8115
## Proportion of Variance 0.2735 0.2267 0.1938 0.1742 0.1317
## Cumulative Proportion  0.2735 0.5002 0.6940 0.8683 1.0000

fviz_eig(pca_bab)

fviz_pca_var(pca_bab)

###(1) 각 PC축에 가깝게 평행을 이루는 변수가 해당 PC에 영향을 가장 많이 주는 변수.

: PC 1에 가장 영향을 많이 주는 변수는 host_listing 

###(2) 각 빨간선의 길이는 원변수의 분산을 표현, 길이가 길수록 분산이 길다.

: 그 다음으로 PC1 에 영향을 주는 변수는 mimum_nights 로 볼수 있다. 

###(3) 각 빨간선이 가까울수록 서로 상관관계가 있다. (반대로 서로 거리가 멀수록 상관관계가 적다.)

: listing_count, price, mimum_night는 상관관계가 높다고 판단됨 

PCA - Crime

범죄에 가장 연관성이 있는 변수들을 파악해보기

USArrest 데이터 세트 활용

data("USArrests")



round(stat.desc(USArrests), digits = 3)->df

DT::datatable(df)

boxplot(USArrests)

library(corrplot)
cor(USArrests, method = "pearson") -> corr

corrplot(corr, method = "number") -> corr_image

prcomp(USArrests, scale = T) ->pca_crime

summary(pca_crime)

## Importance of components:
##                           PC1    PC2     PC3     PC4
## Standard deviation     1.5749 0.9949 0.59713 0.41645
## Proportion of Variance 0.6201 0.2474 0.08914 0.04336
## Cumulative Proportion  0.6201 0.8675 0.95664 1.00000

screeplot(pca_crime, type='l')

Variance 가 1 이하로 넘어가지 않는걸 생각하면 2개의 주성분이 유효함

pca_crime$sdev

## [1] 1.5748783 0.9948694 0.5971291 0.4164494

pca_crime$rotation

##                 PC1        PC2        PC3         PC4
## Murder   -0.5358995  0.4181809 -0.3412327  0.64922780
## Assault  -0.5831836  0.1879856 -0.2681484 -0.74340748
## UrbanPop -0.2781909 -0.8728062 -0.3780158  0.13387773
## Rape     -0.5434321 -0.1673186  0.8177779  0.08902432

fviz_eig(pca_crime)

fviz_pca(pca_crime, 
         labelsize=3)+theme_bw()

fviz_pca_var(pca_crime)

fviz_contrib(pca_crime, choice = "var", axes=1)

fviz_pca_var(pca_crime, col.var = "contrib",
             gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"))

K-Means

K-mean 두가지 예제

Wholesale customers

setwd("C:/Users/Administrator/Desktop/R Analysis/Fast Campus")

read.csv("Wholesale customers data.csv", header = T, 
         stringsAsFactors = T) -> df

head(df)

##   Channel Region Fresh Milk Grocery Frozen Detergents_Paper Delicassen
## 1       2      3 12669 9656    7561    214             2674       1338
## 2       2      3  7057 9810    9568   1762             3293       1776
## 3       2      3  6353 8808    7684   2405             3516       7844
## 4       1      3 13265 1196    4221   6404              507       1788
## 5       2      3 22615 5410    7198   3915             1777       5185
## 6       2      3  9413 8259    5126    666             1795       1451

Channel 과 Region 변수를 바꿔주기

df$Channel <- as.factor(df$Channel)
df$Region <- as.factor(df$Region)

결측치 확인

colSums(is.na(df))

##          Channel           Region            Fresh             Milk 
##                0                0                0                0 
##          Grocery           Frozen Detergents_Paper       Delicassen 
##                0                0                0                0

기술통계와 분포

stat.desc(df,  basic = F)

##          Channel Region        Fresh         Milk      Grocery       Frozen
## median        NA     NA 8.504000e+03 3.627000e+03 4.755500e+03 1.526000e+03
## mean          NA     NA 1.200030e+04 5.796266e+03 7.951277e+03 3.071932e+03
## SE.mean       NA     NA 6.029377e+02 3.518457e+02 4.530455e+02 2.314375e+02
## CI.mean       NA     NA 1.185003e+03 6.915113e+02 8.904077e+02 4.548631e+02
## var           NA     NA 1.599549e+08 5.446997e+07 9.031010e+07 2.356785e+07
## std.dev       NA     NA 1.264733e+04 7.380377e+03 9.503163e+03 4.854673e+03
## coef.var      NA     NA 1.053918e+00 1.273299e+00 1.195174e+00 1.580332e+00
##          Detergents_Paper   Delicassen
## median       8.165000e+02 9.655000e+02
## mean         2.881493e+03 1.524870e+03
## SE.mean      2.272985e+02 1.344433e+02
## CI.mean      4.467286e+02 2.642325e+02
## var          2.273244e+07 7.952997e+06
## std.dev      4.767854e+03 2.820106e+03
## coef.var     1.654647e+00 1.849407e+00

summary(df)

##  Channel Region      Fresh             Milk          Grocery     
##  1:298   1: 77   Min.   :     3   Min.   :   55   Min.   :    3  
##  2:142   2: 47   1st Qu.:  3128   1st Qu.: 1533   1st Qu.: 2153  
##          3:316   Median :  8504   Median : 3627   Median : 4756  
##                  Mean   : 12000   Mean   : 5796   Mean   : 7951  
##                  3rd Qu.: 16934   3rd Qu.: 7190   3rd Qu.:10656  
##                  Max.   :112151   Max.   :73498   Max.   :92780  
##      Frozen        Detergents_Paper    Delicassen     
##  Min.   :   25.0   Min.   :    3.0   Min.   :    3.0  
##  1st Qu.:  742.2   1st Qu.:  256.8   1st Qu.:  408.2  
##  Median : 1526.0   Median :  816.5   Median :  965.5  
##  Mean   : 3071.9   Mean   : 2881.5   Mean   : 1524.9  
##  3rd Qu.: 3554.2   3rd Qu.: 3922.0   3rd Qu.: 1820.2  
##  Max.   :60869.0   Max.   :40827.0   Max.   :47943.0

boxplot(df[,3:ncol(df)]) #ncol : 전체 열의 데이터를 가져오기 

K-mean 는 이상치의 영향을 많이 받기 때문에, 이상치를 제거해주는 것이 좋다.

temp <-NULL 


for(i in 3:ncol(df)) {
    temp <-rbind(temp, df[order(df[,i], 
                                decreasing = T),] 
    %>% slice(1:5))
}


temp %>% 
  arrange(Fresh) %>% 
  head()

##   Channel Region Fresh  Milk Grocery Frozen Detergents_Paper Delicassen
## 1       2      3    85 20959   45828     36            24231       1423
## 2       2      3    85 20959   45828     36            24231       1423
## 3       2      2  8565  4980   67298    131            38102       1215
## 4       2      2  8565  4980   67298    131            38102       1215
## 5       1      3 11314  3090    2062  35009               71       2698
## 6       2      3 16117 46197   92780   1026            40827       2944

#복원 추출 방식으로 중복이 발생한다

중복을 제거하기 distinct 함수

temp <-distinct(temp)

anti_join(df, temp) -> df.rm.outlier 

# df에서 temp 내용을 지워라 

par(mfrow = c(1,2))
boxplot(df[,3:ncol(df)])
boxplot(df.rm.outlier[,3:ncol(df)])

dev.off()

## null device 
##           1

1.K 군집 개수 설정 (Elbow Method)

WSS 의 최소 지점 : 5 개

library(factoextra)
set.seed(1234)

fviz_nbclust(df.rm.outlier[, 3:ncol(df.rm.outlier)], 
             kmeans, method = "wss", k.max = 15)+ 
  theme_minimal()

2.K 군집 개수 설정 (Silloutte Method)

K 는 3개

set.seed(1234)

fviz_nbclust(df.rm.outlier[, 3:ncol(df.rm.outlier)], 
             kmeans, method = "silhouette", k.max = 15)+ 
  theme_minimal()

구매 데이터의 고객 클러스터링이기 때문에 K 가 5개 인게 나을것 같다

3.K-Means Modelling

kmeans(df.rm.outlier[,3:ncol(df.rm.outlier)], 
                     center = 5, 
                     iter.max= 1000) -> df.kmeans

#iter.max : 군집화 후, 재 군집화 과정에서 몇번 반복 시킬 것인가 

df.kmeans

## K-means clustering with 5 clusters of sizes 179, 42, 72, 110, 18
## 
## Cluster means:
##       Fresh      Milk   Grocery   Frozen Detergents_Paper Delicassen
## 1  4267.933  3751.480  4672.950 2211.313        1550.4469   1036.006
## 2 25332.000  5603.548  7160.024 4144.667        1449.2381   2053.333
## 3  5152.250 12536.694 19616.472 1644.014        8794.1389   1696.653
## 4 14527.509  2606.064  3503.873 3202.073         804.8091   1037.882
## 5 40558.056  3113.444  3814.333 2974.833         684.2778   1271.333
## 
## Clustering vector:
##   [1] 4 1 1 4 2 1 4 1 1 3 1 4 2 2 2 4 1 1 2 1 4 1 2 2 4 4 4 3 5 2 1 4 2 1 1 2 3
##  [38] 3 2 4 3 3 1 3 3 4 3 1 1 5 3 2 1 3 3 4 1 1 1 3 1 1 2 1 1 4 1 2 1 4 1 3 4 1
##  [75] 1 3 1 4 4 1 2 4 4 3 3 1 1 1 1 4 3 3 1 4 4 1 3 1 3 4 3 4 4 4 4 4 1 4 1 4 1
## [112] 4 4 5 4 2 1 5 1 1 4 4 1 1 1 1 4 1 4 2 5 4 4 3 1 1 1 5 4 1 4 1 1 3 3 4 1 3
## [149] 1 4 4 3 1 3 1 1 1 1 3 3 1 3 1 1 5 4 4 1 4 1 1 1 1 1 3 3 4 4 1 3 1 4 1 4 4
## [186] 3 3 2 1 1 3 1 1 1 3 4 3 1 1 1 3 3 4 3 1 4 1 1 1 1 4 2 1 1 1 4 1 2 1 4 1 1
## [223] 4 1 5 2 2 4 4 1 3 1 4 4 1 1 3 1 2 1 5 4 1 5 1 1 2 1 3 3 3 4 3 4 1 1 1 5 1
## [260] 1 2 4 4 4 1 4 5 2 5 1 4 4 5 1 1 1 3 2 1 4 1 1 1 4 3 1 3 3 1 3 4 1 3 1 2 3
## [297] 4 4 3 1 1 4 3 1 1 4 4 2 1 1 4 1 4 3 2 4 2 4 4 1 1 1 1 1 3 1 1 3 2 1 3 1 3
## [334] 1 3 4 1 4 3 1 1 4 1 1 1 1 1 4 1 2 1 5 4 1 4 1 1 3 5 1 1 2 4 5 1 3 4 1 4 4
## [371] 4 1 1 1 2 4 4 1 4 4 4 1 2 2 2 4 1 2 3 1 1 1 1 1 1 1 1 3 1 3 1 3 4 2 4 4 4
## [408] 3 2 1 1 1 1 4 1 4 2 5 3 4 1
## 
## Within cluster sum of squares by cluster:
## [1] 7488224454 2823135964 9143410363 3900150510  861057236
##  (between_SS / total_SS =  70.9 %)
## 
## Available components:
## 
## [1] "cluster"      "centers"      "totss"        "withinss"     "tot.withinss"
## [6] "betweenss"    "size"         "iter"         "ifault"

4.시각화

barplot(t(df.kmeans$centers), beside=TRUE, col = 1:6)
legend("topleft", colnames(df[,3:8]), fill = 1:6, cex = 0.5)

df.rm.outlier$cluster <- df.kmeans$cluster

head(df.rm.outlier)

##   Channel Region Fresh Milk Grocery Frozen Detergents_Paper Delicassen cluster
## 1       2      3 12669 9656    7561    214             2674       1338       4
## 2       2      3  7057 9810    9568   1762             3293       1776       1
## 3       2      3  6353 8808    7684   2405             3516       7844       1
## 4       1      3 13265 1196    4221   6404              507       1788       4
## 5       2      3 22615 5410    7198   3915             1777       5185       2
## 6       2      3  9413 8259    5126    666             1795       1451       1

Arbnb

setwd("C:/Users/Administrator/Desktop/R Analysis/Business R Aanlysis source")

read.csv("AB_NYC_2019.csv") -> df

str(df)

## 'data.frame':    48895 obs. of  16 variables:
##  $ id                            : int  2539 2595 3647 3831 5022 5099 5121 5178 5203 5238 ...
##  $ name                          : chr  "Clean & quiet apt home by the park" "Skylit Midtown Castle" "THE VILLAGE OF HARLEM....NEW YORK !" "Cozy Entire Floor of Brownstone" ...
##  $ host_id                       : int  2787 2845 4632 4869 7192 7322 7356 8967 7490 7549 ...
##  $ host_name                     : chr  "John" "Jennifer" "Elisabeth" "LisaRoxanne" ...
##  $ neighbourhood_group           : chr  "Brooklyn" "Manhattan" "Manhattan" "Brooklyn" ...
##  $ neighbourhood                 : chr  "Kensington" "Midtown" "Harlem" "Clinton Hill" ...
##  $ latitude                      : num  40.6 40.8 40.8 40.7 40.8 ...
##  $ longitude                     : num  -74 -74 -73.9 -74 -73.9 ...
##  $ room_type                     : chr  "Private room" "Entire home/apt" "Private room" "Entire home/apt" ...
##  $ price                         : int  149 225 150 89 80 200 60 79 79 150 ...
##  $ minimum_nights                : int  1 1 3 1 10 3 45 2 2 1 ...
##  $ number_of_reviews             : int  9 45 0 270 9 74 49 430 118 160 ...
##  $ last_review                   : chr  "2018-10-19" "2019-05-21" "" "2019-07-05" ...
##  $ reviews_per_month             : num  0.21 0.38 NA 4.64 0.1 0.59 0.4 3.47 0.99 1.33 ...
##  $ calculated_host_listings_count: int  6 2 1 1 1 1 1 1 1 4 ...
##  $ availability_365              : int  365 355 365 194 0 129 0 220 0 188 ...

df %>%  
  select(-reviews_per_month, -last_review, -latitude, -longitude) %>%  
  mutate(name = as.character(name), 
         id = as.character(id), 
         host_id= as.character(host_id), 
         hots_name = as.character(host_name),
         price = as.numeric(price)) -> df_2


df_num <-df_2 %>% 
  select_if(is.numeric)

boxplot(df_num)    

#--------------------------------------------
# price 부분의 outlier 삭제 
#---------------------------------------------



df_num %>% 
  filter(availability_365 == 0) -> temp

anti_join(df_num, temp) -> df_num_outlier

df_num_outlier %>% 
  filter(price >= 9000 ) -> temp_2


anti_join(df_num_outlier, temp_2) -> df_2

boxplot(df_2$price)

#--------------------------------------------
# Scaling 
#---------------------------------------------


df_num_scale <- as.data.frame(scale(df_2))

kmeans(df_num_scale,7) -> df_km

fviz_cluster(df_km, 
             data=df_num_scale)+
  theme_minimal()

#--------------------------------------------
# K 찾기 - manual 
#---------------------------------------------

# Determine K
wss <- function(data, maxCluster = 20) {
  # Initialize within sum of squares
  SSw <- (nrow(data) - 1) * sum(apply(data, 2, var))
  SSw <- vector()
  for (i in 2:maxCluster) {
    SSw[i] <- sum(kmeans(data, centers = i)$withinss)
  }
  plot(1:maxCluster, SSw, type = "o", xlab = "Number of Clusters", ylab = "Within groups sum of squares", pch=19)
}

wss(df_num_scale)  ##8개 

kmeans(df_num_scale, 8) -> df_km8


fviz_cluster(df_km8, 
             data=df_num_scale)

df_km8$cluster -> df_2$cluster


barplot(t(df_km8$centers), beside=TRUE, col = 1:8)
legend("topleft", colnames(df_2[,1:4]), fill=1:8, cex=0.5)