ANOVA TEST : 探討不同運費下船隻駁油量是否有顯著差異

讀入檔案

把運費轉成類別變數

Tb <- read.csv("T.csv")
colnames(Tb) <- c("Date","Spinfo","KL","Bill","trans_fee","TotalP","O_price")
Tb$trans_fee <- as.factor(Tb$trans_fee) #5個Levels

變異數同值?

Levene檢定

library(car)

## Warning: package 'car' was built under R version 3.4.4

## Loading required package: carData

## Warning: package 'carData' was built under R version 3.4.4

leveneTest(Tb$KL, Tb$trans_fee, center=mean)

## Levene's Test for Homogeneity of Variance (center = mean)
##        Df F value Pr(>F)
## group   4  0.6431 0.6324
##       187

 #Levene檢定，leveneTest結果p-value=0.6324 > 0.05，所以無法證明兩變異數不相同(指變異數相同):不顯著

result1 <- aov(Tb$KL ~ Tb$trans_fee)
summary(result1)  #ANOVA不顯著，不拒絕虛無假設

##               Df   Sum Sq Mean Sq F value Pr(>F)
## Tb$trans_fee   4   129690   32423    0.23  0.921
## Residuals    187 26304851  140668

TukeyHSD(result1)

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = Tb$KL ~ Tb$trans_fee)
## 
## $`Tb$trans_fee`
##               diff       lwr      upr     p adj
## 188-168 -65.431372 -402.4422 271.5795 0.9836293
## 407-168 -69.820250 -559.8736 420.2331 0.9949500
## 414-168 -37.311500 -527.3649 452.7419 0.9995662
## 489-168  33.518000 -428.5088 495.5448 0.9996417
## 407-188  -4.388878 -378.9018 370.1240 0.9999998
## 414-188  28.119872 -346.3930 402.6328 0.9995894
## 489-188  98.949372 -238.0615 435.9602 0.9276815
## 414-407  32.508750 -484.0529 549.0704 0.9997961
## 489-407 103.338250 -386.7151 593.3916 0.9777567
## 489-414  70.829500 -419.2239 560.8829 0.9946629

diff 表示差距:運費188和168的KL差距是65.4313

lwr , upr 不包含0就會顯著

我們發現不同的運費定價對於船舶駁油的量並沒有顯著差異。