Dataset and library

library("foreign", lib.loc="~/R/win-library/3.2")

## Warning: package 'foreign' was built under R version 3.2.5

library(WRS2)

## Warning: package 'WRS2' was built under R version 3.2.5

##data for DEE
df=read.csv("C:/Users/BINH THANG/Dropbox/R - Learning/drbinh/2017/slieubieudo.csv")

##data for type 2 EEL (graph 2 cua anh)

df1=read.csv("C:/Users/BINH THANG/Dropbox/R - Learning/drbinh/2017/bieudo2.csv")

pre

formula1="Preb~nhom1"
yuen(formula=formula1, data=df,tr = 0.2)

## Call:
## yuen(formula = formula1, data = df, tr = 0.2)
## 
## Test statistic: 3.2208 (df = 9.82), p-value = 0.00937
## 
## Trimmed mean difference:  12.51978 
## 95 percent confidence interval:
## 3.8373     21.2023

yuenbt(formula1, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula1, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 3.1376 (df = NA), p-value = 0.016
## 
## Trimmed mean difference:  12.51978 
## 95 percent confidence interval:
## 3.3817     21.6578

30days

formula2="e30days~nhom1"
yuen(formula=formula2, data=df,tr = 0.2)

## Call:
## yuen(formula = formula2, data = df, tr = 0.2)
## 
## Test statistic: 2.7358 (df = 9.58), p-value = 0.02177
## 
## Trimmed mean difference:  11.65476 
## 95 percent confidence interval:
## 2.1065     21.2031

yuenbt(formula2, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula2, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 2.6841 (df = NA), p-value = 0.024
## 
## Trimmed mean difference:  11.65476 
## 95 percent confidence interval:
## 2.0994     21.2101

6 mnonths

formula4="e6months~nhom1"
yuen(formula=formula4, data=df,tr = 0.2)

## Call:
## yuen(formula = formula4, data = df, tr = 0.2)
## 
## Test statistic: 2.6951 (df = 13.1), p-value = 0.01826
## 
## Trimmed mean difference:  12.41714 
## 95 percent confidence interval:
## 2.4716     22.3626

yuenbt(formula4, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula4, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 2.2934 (df = NA), p-value = 0.023
## 
## Trimmed mean difference:  12.41714 
## 95 percent confidence interval:
## 2.3435     22.4908

1year

formula5="e1year~nhom1"
yuen(formula=formula5, data=df,tr = 0.2)

## Call:
## yuen(formula = formula5, data = df, tr = 0.2)
## 
## Test statistic: 2.4129 (df = 3.62), p-value = 0.08002
## 
## Trimmed mean difference:  18.825 
## 95 percent confidence interval:
## -3.7576     41.4076

yuenbt(formula5, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula5, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 1.491 (df = NA), p-value = 0.206
## 
## Trimmed mean difference:  18.825 
## 95 percent confidence interval:
## -18.6409     56.2909

2year

formula6="e2years~nhom1"
yuen(formula=formula6, data=df,tr = 0.2)

## Call:
## yuen(formula = formula6, data = df, tr = 0.2)
## 
## Test statistic: 0.4024 (df = 9.52), p-value = 0.6963
## 
## Trimmed mean difference:  2.85 
## 95 percent confidence interval:
## -13.0412     18.7412

yuenbt(formula6, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula6, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 0.2879 (df = NA), p-value = 0.691
## 
## Trimmed mean difference:  2.85 
## 95 percent confidence interval:
## -15.3851     21.0851

3 year

formula7="e3years~nhom1"
yuen(formula=formula7, data=df,tr = 0.2)

## Call:
## yuen(formula = formula7, data = df, tr = 0.2)
## 
## Test statistic: 0.6719 (df = 4.97), p-value = 0.53164
## 
## Trimmed mean difference:  6.20833 
## 95 percent confidence interval:
## -17.5936     30.0102

yuenbt(formula7, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula7, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 0.6757 (df = NA), p-value = 0.538
## 
## Trimmed mean difference:  6.20833 
## 95 percent confidence interval:
## -17.4599     29.8766

4 year

formula8="e4years~nhom1"
yuen(formula=formula8, data=df,tr = 0.2)

## Call:
## yuen(formula = formula8, data = df, tr = 0.2)
## 
## Test statistic: 2.7718 (df = 3.12), p-value = 0.06642
## 
## Trimmed mean difference:  23.08667 
## 95 percent confidence interval:
## -2.8476     49.021

yuenbt(formula8, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula8, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 2.8619 (df = NA), p-value = 0.049
## 
## Trimmed mean difference:  23.08667 
## 95 percent confidence interval:
## 0.1908     45.9825

5 year

formula9="e5years~nhom1"
yuen(formula=formula9, data=df,tr = 0.2)

## Call:
## yuen(formula = formula9, data = df, tr = 0.2)
## 
## Test statistic: 0.8589 (df = 5.7), p-value = 0.42506
## 
## Trimmed mean difference:  12.72667 
## 95 percent confidence interval:
## -24.0019     49.4552

yuenbt(formula9, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula9, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 0.6126 (df = NA), p-value = 0.42
## 
## Trimmed mean difference:  12.72667 
## 95 percent confidence interval:
## -28.6923     54.1456

6 year

formula10="e6years~nhom1"
yuen(formula=formula10, data=df,tr = 0.2)

## Call:
## yuen(formula = formula10, data = df, tr = 0.2)
## 
## Test statistic: 1.8541 (df = 3.85), p-value = 0.14016
## 
## Trimmed mean difference:  36.3 
## 95 percent confidence interval:
## -18.9162     91.5162

yuenbt(formula10, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula10, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 1.1124 (df = NA), p-value = 0.155
## 
## Trimmed mean difference:  36.3 
## 95 percent confidence interval:
## -26.6172     99.2172

7 year

formula11="e7years~nhom1"
yuen(formula=formula11, data=df,tr = 0.2)

## Call:
## yuen(formula = formula11, data = df, tr = 0.2)
## 
## Test statistic: 0.5752 (df = 4.03), p-value = 0.59576
## 
## Trimmed mean difference:  11.59167 
## 95 percent confidence interval:
## -44.1874     67.3707

yuenbt(formula11, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula11, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: 0.3451 (df = NA), p-value = 0.583
## 
## Trimmed mean difference:  11.59167 
## 95 percent confidence interval:
## -74.4915     97.6748

8 year

formula12="e8years~nhom1"
yuen(formula=formula12, data=df,tr = 0.2)

## Call:
## yuen(formula = formula12, data = df, tr = 0.2)
## 
## Test statistic: NA (df = NA), p-value = NA
## 
## Trimmed mean difference:  32 
## 95 percent confidence interval:
## NA     NA

yuenbt(formula12, data=df, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula12, data = df, tr = 0.2, nboot = 1000)
## 
## Test statistic: NA (df = NA), p-value = 0
## 
## Trimmed mean difference:  32 
## 95 percent confidence interval:
## NA     NA

11 year

#formula="e11y~nhom1"
#yuen(formula=formula, data=df,tr = 0.2)
#yuenbt(formula, data=df, tr = 0.2, nboot = 1000)

P-value for 2nd

Dataset and library

df1=read.csv("C:/Users/BINH THANG/Dropbox/R - Learning/drbinh/2017/bieudo2.csv")

pre

formula1="Preb~nhom1"
yuen(formula=formula1, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula1, data = df1, tr = 0.2)
## 
## Test statistic: 2.3912 (df = 2.25), p-value = 0.12514
## 
## Trimmed mean difference:  14.53333 
## 95 percent confidence interval:
## -9.0054     38.0721

yuenbt(formula1, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula1, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: 2.5934 (df = NA), p-value = 0.282
## 
## Trimmed mean difference:  14.53333 
## 95 percent confidence interval:
## -19.67     48.7366

30days

formula2="e30days~nhom1"
yuen(formula=formula2, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula2, data = df1, tr = 0.2)
## 
## Test statistic: 2.7081 (df = 2.5), p-value = 0.08947
## 
## Trimmed mean difference:  13.80333 
## 95 percent confidence interval:
## -4.4301     32.0368

yuenbt(formula2, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula2, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: 2.9279 (df = NA), p-value = 0.132
## 
## Trimmed mean difference:  13.80333 
## 95 percent confidence interval:
## -5.7801     33.3867

6 mnonths

formula4="e6months~nhom1"
yuen(formula=formula4, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula4, data = df1, tr = 0.2)
## 
## Test statistic: 2.919 (df = 3.36), p-value = 0.05361
## 
## Trimmed mean difference:  12.8875 
## 95 percent confidence interval:
## -0.3515     26.1265

yuenbt(formula4, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula4, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: 3.0502 (df = NA), p-value = 0.043
## 
## Trimmed mean difference:  12.8875 
## 95 percent confidence interval:
## 0.747     25.028

1year

formula5="e1year~nhom1"
yuen(formula=formula5, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula5, data = df1, tr = 0.2)
## 
## Test statistic: 4.4309 (df = 1.76), p-value = 0.05965
## 
## Trimmed mean difference:  14.84444 
## 95 percent confidence interval:
## -1.6649     31.3538

yuenbt(formula5, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula5, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: 2.8585 (df = NA), p-value = 0.064
## 
## Trimmed mean difference:  14.84444 
## 95 percent confidence interval:
## -1.6705     31.3594

2year

formula6="e2years~nhom1"
yuen(formula=formula6, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula6, data = df1, tr = 0.2)
## 
## Test statistic: 0.9448 (df = 3.45), p-value = 0.40608
## 
## Trimmed mean difference:  4.56667 
## 95 percent confidence interval:
## -9.7363     18.8696

yuenbt(formula6, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula6, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: 0.6226 (df = NA), p-value = 0.446
## 
## Trimmed mean difference:  4.56667 
## 95 percent confidence interval:
## -21.7489     30.8822

3 year

formula7="e3years~nhom1"
yuen(formula=formula7, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula7, data = df1, tr = 0.2)
## 
## Test statistic: 0.8947 (df = 3.46), p-value = 0.42886
## 
## Trimmed mean difference:  8.475 
## 95 percent confidence interval:
## -19.5392     36.4892

yuenbt(formula7, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula7, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: 0.5368 (df = NA), p-value = 0.395
## 
## Trimmed mean difference:  8.475 
## 95 percent confidence interval:
## -27.8464     44.7964

4 year

formula8="e4years~nhom1"
yuen(formula=formula8, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula8, data = df1, tr = 0.2)
## 
## Test statistic: NA (df = NA), p-value = NA
## 
## Trimmed mean difference:  13.425 
## 95 percent confidence interval:
## NA     NA

yuenbt(formula8, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula8, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: NA (df = NA), p-value = 0
## 
## Trimmed mean difference:  13.425 
## 95 percent confidence interval:
## NA     NA

5 year

formula9="e5years~nhom1"
yuen(formula=formula9, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula9, data = df1, tr = 0.2)
## 
## Test statistic: 0.5126 (df = 3.49), p-value = 0.63892
## 
## Trimmed mean difference:  5.94167 
## 95 percent confidence interval:
## -28.175     40.0583

yuenbt(formula9, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula9, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: 0.3076 (df = NA), p-value = 0.618
## 
## Trimmed mean difference:  5.94167 
## 95 percent confidence interval:
## -26.9952     38.8785

6 year

formula10="e6years~nhom1"
yuen(formula=formula10, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula10, data = df1, tr = 0.2)
## 
## Test statistic: NA (df = NA), p-value = NA
## 
## Trimmed mean difference:  32.25 
## 95 percent confidence interval:
## NA     NA

yuenbt(formula10, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula10, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: NA (df = NA), p-value = 0
## 
## Trimmed mean difference:  32.25 
## 95 percent confidence interval:
## NA     NA

7 year

formula11="e7years~nhom1"
yuen(formula=formula11, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula11, data = df1, tr = 0.2)
## 
## Test statistic: NA (df = NA), p-value = NA
## 
## Trimmed mean difference:  21.225 
## 95 percent confidence interval:
## NA     NA

yuenbt(formula11, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula11, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: NA (df = NA), p-value = 0
## 
## Trimmed mean difference:  21.225 
## 95 percent confidence interval:
## NA     NA

8 year

formula12="e8years~nhom1"
yuen(formula=formula12, data=df1,tr = 0.2)

## Call:
## yuen(formula = formula12, data = df1, tr = 0.2)
## 
## Test statistic: NA (df = NA), p-value = NA
## 
## Trimmed mean difference:  32 
## 95 percent confidence interval:
## NA     NA

yuenbt(formula12, data=df1, tr = 0.2, nboot = 1000)

## Call:
## yuenbt(formula = formula12, data = df1, tr = 0.2, nboot = 1000)
## 
## Test statistic: NA (df = NA), p-value = 0
## 
## Trimmed mean difference:  32 
## 95 percent confidence interval:
## NA     NA

p for trend - bieu do cua anh

p for bieu do1

reshape data 1 -

library(reshape2)

## Warning: package 'reshape2' was built under R version 3.2.5

names(df)

##  [1] "stt"      "ID"       "Preb"     "e30days"  "e6months" "e1year"  
##  [7] "e2years"  "e3years"  "e4years"  "e5years"  "e6years"  "e7years" 
## [13] "e8years"  "e11y"     "gd"       "nhom1"

d<- melt(df, id.vars = c("ID", "nhom1"), measure.vars = c( "Preb","e30days",  "e6months", "e1year","e2years" , "e3years" , "e4years" , "e5years"  ,"e6years" , "e7years" ,"e8years" ),variable.name = "time")


d$time1=as.numeric(d$time)

d=na.omit(d)

$p for trend using robus regression

library(MASS)

#p chung cho ca 2 nhom

a1=rlm(d$value~ d$time1+d$nhom1)
summary(a1)

## 
## Call: rlm(formula = d$value ~ d$time1 + d$nhom1)
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -35.0579  -7.7703  -0.4385   7.5199  57.3713 
## 
## Coefficients:
##             Value    Std. Error t value 
## (Intercept)  71.6606   2.4830    28.8610
## d$time1       0.1097   0.3903     0.2812
## d$nhom1     -14.1141   2.1852    -6.4589
## 
## Residual standard error: 11.51 on 170 degrees of freedom

a1=rlm(d$value~ d$time1+d$nhom1)
summary(a1)

## 
## Call: rlm(formula = d$value ~ d$time1 + d$nhom1)
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -35.0579  -7.7703  -0.4385   7.5199  57.3713 
## 
## Coefficients:
##             Value    Std. Error t value 
## (Intercept)  71.6606   2.4830    28.8610
## d$time1       0.1097   0.3903     0.2812
## d$nhom1     -14.1141   2.1852    -6.4589
## 
## Residual standard error: 11.51 on 170 degrees of freedom

#p trend cho nhom 11 cases (late?)

d1=subset(d, nhom1==0)

a2=lm(value~ time1, data=d1)
summary(a2)

## 
## Call:
## lm(formula = value ~ time1, data = d1)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -39.964 -10.424  -0.981   8.544  54.496 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  69.7972     4.2304  16.499   <2e-16 ***
## time1         0.7867     0.7728   1.018    0.313    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.41 on 62 degrees of freedom
## Multiple R-squared:  0.01644,    Adjusted R-squared:  0.0005759 
## F-statistic: 1.036 on 1 and 62 DF,  p-value: 0.3126

#p trend cho nhom 22 cases (early)
d2=subset(d, nhom1==1)
a3=lm(value~ time1, data=d2)
summary(a3)

## 
## Call:
## lm(formula = value ~ time1, data = d2)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -24.392  -7.171  -0.879   7.021  34.008 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  59.2664     2.0960  28.276   <2e-16 ***
## time1        -0.2974     0.4603  -0.646     0.52    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 11.94 on 107 degrees of freedom
## Multiple R-squared:  0.003887,   Adjusted R-squared:  -0.005422 
## F-statistic: 0.4175 on 1 and 107 DF,  p-value: 0.5195

#Overall, it's shown p=0.634, which indicated no significant trend increasing or decreasing. However, we found statistical significantly diffirence in 2 groups (p=9.47e-10)
#In particualr, no significant was observed in both groups (p=0.0.313 , p=0.52, respectively)

p for bieu do2

reshape data 2 -

d00<- melt(df1, id.vars = c("stt", "nhom1"), measure.vars = c( "Preb","e30days",  "e6months", "e1year","e2years" , "e3years" , "e4years" , "e5years"  ,"e6years" , "e7years" ,"e8years" ),variable.name = "time")
d00$time1=as.numeric(d00$time)
d00=na.omit(d00)

p for trend using linear regression and robus regression (confirmed)

a=rlm(d00$value~ d00$time1+d00$nhom1, data=d00)
summary(a)

## 
## Call: rlm(formula = d00$value ~ d00$time1 + d00$nhom1, data = d00)
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -21.893  -5.554   0.151   5.549  36.507 
## 
## Coefficients:
##             Value    Std. Error t value 
## (Intercept)  70.9101   2.2703    31.2340
## d00$time1    -0.3458   0.3259    -1.0610
## d00$nhom1   -13.6592   2.0248    -6.7460
## 
## Residual standard error: 8.266 on 115 degrees of freedom

a=lm(d00$value~ d00$time1+d00$nhom1)
summary(a)

## 
## Call:
## lm(formula = d00$value ~ d00$time1 + d00$nhom1)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -22.498  -5.427   0.145   5.907  35.902 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  69.7929     2.4636  28.329  < 2e-16 ***
## d00$time1    -0.2730     0.3537  -0.772    0.442    
## d00$nhom1   -12.6648     2.1972  -5.764 7.01e-08 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 10.35 on 115 degrees of freedom
## Multiple R-squared:  0.2245, Adjusted R-squared:  0.2111 
## F-statistic: 16.65 on 2 and 115 DF,  p-value: 4.461e-07

#it's shown p=0.44, which indicated no significant trend increasing or decreasing. However, we found statistical significantly diffirence in 2 groups (p=7.01e-08). However, ... significant dif between 2 gorups

#p trend cho nhom 11 cases (late?)

d2=subset(d00, nhom1==0)

a2=lm(value~ time1, data=d00)
summary(a2)

## 
## Call:
## lm(formula = value ~ time1, data = d00)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -26.759  -7.771  -0.946   6.208  31.641 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 59.63395    1.94583  30.647   <2e-16 ***
## time1       -0.09751    0.39832  -0.245    0.807    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 11.7 on 116 degrees of freedom
## Multiple R-squared:  0.0005164,  Adjusted R-squared:  -0.0081 
## F-statistic: 0.05993 on 1 and 116 DF,  p-value: 0.807

#p trend cho nhom 22 cases (early)
d2=subset(d00, nhom1==1)
a3=lm(value~ time1, data=d2)
summary(a3)

## 
## Call:
## lm(formula = value ~ time1, data = d2)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -21.827  -5.450   0.047   4.521  36.573 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  57.5626     2.0657  27.866   <2e-16 ***
## time1        -0.3835     0.4377  -0.876    0.383    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 10.72 on 86 degrees of freedom
## Multiple R-squared:  0.008849,   Adjusted R-squared:  -0.002676 
## F-statistic: 0.7679 on 1 and 86 DF,  p-value: 0.3833

#Conclusion: no significant was observed in both groups (p=0.807 , p=0.383respectively

p-value

Dr B

21 April 2017

Dataset and library

pre

30days

6 mnonths

1year

2year

3 year

4 year

5 year

6 year

7 year

8 year

11 year

Dataset and library

pre

30days

6 mnonths

1year

2year

3 year

4 year

5 year

6 year

7 year

8 year

p for trend - bieu do cua anh

p for bieu do1

reshape data 1 -

p for bieu do2

reshape data 2 -

p for trend using linear regression and robus regression (confirmed)