BMA - Linear regression and
Thang, TB
22 March 2017
BMA
library(foreign)## Warning: package 'foreign' was built under R version 3.2.5r=read.dta("C:/Users/BINH THANG/Dropbox/Korea/STudy/Thesis/data management/DataR/dataR3.dta")
r1 <- subset(r, cost_inc<180001)
r1$logC=log10(r1$cost_inc)
attach(r1)
newdata2=r1
newdata2$c1=as.factor(newdata2$c1)
newdata2$h1=as.factor(newdata2$h1)
newdata2$label1=as.factor(newdata2$label1)
newdata2$policy=as.factor(newdata2$policy)
newdata2$educ2=as.factor(newdata2$educ2)
newdata2$ter_in=as.factor(newdata2$ter_in)
newdata2$age_group=as.factor(newdata2$age_group)
newdata2$d1a=as.factor(newdata2$d1a)
newdata2$selfhealth=as.factor(newdata2$selfhealth)
newdata2$b18a=as.factor(newdata2$b18a)
newdata2$b16a=as.factor(newdata2$b16a)
newdata2$b6a=as.factor(newdata2$b6a)
newdata2$smostt=as.factor(newdata2$smostt)
newdata2$reasons1=as.factor(newdata2$reasons1)
newdata2$ter_in=as.factor(newdata2$ter_in)
newdata2$group_age1=as.factor(newdata2$group_age1)
newdata2$policy_a=as.factor(newdata2$policy_a)
attach(newdata2 )## The following objects are masked from r1:
##
## a1, advice, advice1, ag, age_group, anticam, anticam2, b1,
## b10, b10a1, b10a2, b10a3, b10a4, b10a5, b10a6, b10a7, b11,
## b11a, b11a2, b11a3, b12, b12a, b13, b14, b15, b16, b16a, b17,
## b18, b18a, b1a, b2, b2a1, b3, b4, b5, b5a, b6, b6a, b6a1,
## b6a2, b6a3, b6a4, b6a5, b6a6, b6a7, b6a7a, b7, b8, b9, br,
## branch, branch1, branch2, branch3, c1, c2, c23a, c3, c4, c5,
## c6, c7, c8, c9, COST, cost_inc, cost1, costincrease, ct, d1,
## d10, d11, d1a, d2, d3, d3a, d4, d5, d6, d7, d8, d9, Decision,
## e1, e2, edu, educ1, educ2, f1, ghi1, ghi2, ghiro2,
## giadinhkoUH, group_age, group_age1, h1, h10a, h10a1, h10a10,
## h10a11, h10a11a, h10a2, h10a3, h10a4, h10a5, h10a6, h10a7,
## h10a8, h10a9, h12, h12a, h12a_1, h12log, h13, h2, h3, h4, h5,
## h6, h7, h8, h9, ha000, itc1, itc2, l1, l2, l3, l4, l5, l6,
## label1, label2, logC, moneyspent, msdt, n01, n02, n03, n05,
## n06, n07, n08, n1, n10, n100, n101, n102, n103, n11, n12, n13,
## n14, n15, n16, n1b, n2, n3, n35, n36, n37, n38, n39,
## n3posterb, n4, n40, n41, n42, n43, n44, n45, n46, n47, n48,
## n49, n5, n50, n51, n52, n53, n54, n55, n56, n57, n58, n59, n6,
## n60, n61, n61a, n62, n62a, n63, n63a, n64, n64a, n65, n65a,
## n66, n66a, n67, n67a, n68, n68a, n7, n77, n78, n7tren, n8,
## n88, n89, n8nha, n9, n97, n98, n99, n9khach, noEnvi2, occup1,
## policy, policy_a, policy2, reasons, reasons1, s1, Screening,
## selfhealth, SH1, smostt, ter_fa1, ter_in, tertile_fa,
## tertile_indi, test, unitsdiffi1, var242, w1, w2, w3, w4t2<-subset(newdata2, select =c(logC,age_group,d1a, educ2, ter_in, selfhealth,h1, b16a, b18a, c1, smostt,policy_a))
t2=na.omit(t2)
yvar = t2[,1]
xvars = t2[,-1]reg = lm(logC ~ ., data=t2)
step(reg, direction="both")## Start: AIC=-1297.44
## logC ~ age_group + d1a + educ2 + ter_in + selfhealth + h1 + b16a +
## b18a + c1 + smostt + policy_a
##
## Df Sum of Sq RSS AIC
## - age_group 4 0.13027 17.430 -1302.3
## - policy_a 2 0.00181 17.301 -1301.4
## - ter_in 2 0.03401 17.333 -1300.6
## - d1a 1 0.00231 17.302 -1299.4
## - c1 1 0.00297 17.302 -1299.4
## - h1 1 0.02488 17.324 -1298.8
## - b16a 1 0.04309 17.342 -1298.4
## - b18a 1 0.05931 17.358 -1298.0
## - smostt 2 0.14747 17.447 -1297.9
## <none> 17.299 -1297.4
## - selfhealth 1 0.38316 17.682 -1290.3
## - educ2 2 0.52021 17.819 -1289.0
##
## Step: AIC=-1302.3
## logC ~ d1a + educ2 + ter_in + selfhealth + h1 + b16a + b18a +
## c1 + smostt + policy_a
##
## Df Sum of Sq RSS AIC
## - policy_a 2 0.00473 17.434 -1306.2
## - ter_in 2 0.05082 17.480 -1305.1
## - c1 1 0.00399 17.433 -1304.2
## - h1 1 0.01942 17.449 -1303.8
## - b16a 1 0.03991 17.469 -1303.3
## - d1a 1 0.04671 17.476 -1303.2
## - smostt 2 0.13815 17.568 -1303.0
## - b18a 1 0.06219 17.492 -1302.8
## <none> 17.430 -1302.3
## + age_group 4 0.13027 17.299 -1297.4
## - selfhealth 1 0.42705 17.857 -1294.2
## - educ2 2 0.71720 18.147 -1289.4
##
## Step: AIC=-1306.18
## logC ~ d1a + educ2 + ter_in + selfhealth + h1 + b16a + b18a +
## c1 + smostt
##
## Df Sum of Sq RSS AIC
## - ter_in 2 0.05295 17.487 -1308.9
## - c1 1 0.00348 17.438 -1308.1
## - h1 1 0.01760 17.452 -1307.8
## - b16a 1 0.04148 17.476 -1307.2
## - d1a 1 0.04960 17.484 -1307.0
## - smostt 2 0.13976 17.574 -1306.8
## - b18a 1 0.06322 17.497 -1306.7
## <none> 17.434 -1306.2
## + policy_a 2 0.00473 17.430 -1302.3
## + age_group 4 0.13319 17.301 -1301.4
## - selfhealth 1 0.42238 17.857 -1298.2
## - educ2 2 0.73103 18.165 -1293.0
##
## Step: AIC=-1308.91
## logC ~ d1a + educ2 + selfhealth + h1 + b16a + b18a + c1 + smostt
##
## Df Sum of Sq RSS AIC
## - c1 1 0.00672 17.494 -1310.8
## - h1 1 0.01947 17.507 -1310.5
## - b16a 1 0.04670 17.534 -1309.8
## - d1a 1 0.05101 17.538 -1309.7
## - smostt 2 0.14065 17.628 -1309.6
## - b18a 1 0.06109 17.548 -1309.5
## <none> 17.487 -1308.9
## + ter_in 2 0.05295 17.434 -1306.2
## + policy_a 2 0.00686 17.480 -1305.1
## + age_group 4 0.15159 17.336 -1304.6
## - selfhealth 1 0.43983 17.927 -1300.5
## - educ2 2 0.76965 18.257 -1294.9
##
## Step: AIC=-1310.75
## logC ~ d1a + educ2 + selfhealth + h1 + b16a + b18a + smostt
##
## Df Sum of Sq RSS AIC
## - h1 1 0.01921 17.513 -1312.3
## - b16a 1 0.04751 17.541 -1311.6
## - d1a 1 0.04778 17.542 -1311.6
## - b18a 1 0.06066 17.555 -1311.3
## <none> 17.494 -1310.8
## + c1 1 0.00672 17.487 -1308.9
## - smostt 2 0.27239 17.766 -1308.3
## + ter_in 2 0.05620 17.438 -1308.1
## + policy_a 2 0.00626 17.488 -1306.9
## + age_group 4 0.15317 17.341 -1306.4
## - selfhealth 1 0.43578 17.930 -1302.4
## - educ2 2 0.76747 18.261 -1296.8
##
## Step: AIC=-1312.29
## logC ~ d1a + educ2 + selfhealth + b16a + b18a + smostt
##
## Df Sum of Sq RSS AIC
## - b16a 1 0.03915 17.552 -1313.4
## - d1a 1 0.05016 17.563 -1313.1
## - b18a 1 0.06565 17.579 -1312.7
## <none> 17.513 -1312.3
## + h1 1 0.01921 17.494 -1310.8
## + c1 1 0.00646 17.507 -1310.5
## - smostt 2 0.26700 17.780 -1310.0
## + ter_in 2 0.05802 17.455 -1309.7
## + policy_a 2 0.00348 17.510 -1308.4
## + age_group 4 0.14605 17.367 -1307.8
## - selfhealth 1 0.42919 17.942 -1304.2
## - educ2 2 0.79381 18.307 -1297.7
##
## Step: AIC=-1313.36
## logC ~ d1a + educ2 + selfhealth + b18a + smostt
##
## Df Sum of Sq RSS AIC
## - d1a 1 0.05296 17.605 -1314.1
## - b18a 1 0.05495 17.607 -1314.0
## <none> 17.552 -1313.4
## + b16a 1 0.03915 17.513 -1312.3
## + h1 1 0.01085 17.541 -1311.6
## + c1 1 0.00725 17.545 -1311.5
## - smostt 2 0.26407 17.816 -1311.1
## + ter_in 2 0.06299 17.489 -1310.9
## + policy_a 2 0.00537 17.547 -1309.5
## + age_group 4 0.14213 17.410 -1308.8
## - selfhealth 1 0.44890 18.001 -1304.8
## - educ2 2 0.77165 18.324 -1299.3
##
## Step: AIC=-1314.09
## logC ~ educ2 + selfhealth + b18a + smostt
##
## Df Sum of Sq RSS AIC
## - b18a 1 0.04480 17.650 -1315.0
## <none> 17.605 -1314.1
## + d1a 1 0.05296 17.552 -1313.4
## + b16a 1 0.04195 17.563 -1313.1
## + h1 1 0.01244 17.593 -1312.4
## - smostt 2 0.24764 17.853 -1312.2
## + c1 1 0.00374 17.602 -1312.2
## + ter_in 2 0.06711 17.538 -1311.7
## + age_group 4 0.19152 17.414 -1310.7
## + policy_a 2 0.00830 17.597 -1310.3
## - selfhealth 1 0.45664 18.062 -1305.4
## - educ2 2 0.96243 18.568 -1295.8
##
## Step: AIC=-1315.03
## logC ~ educ2 + selfhealth + smostt
##
## Df Sum of Sq RSS AIC
## <none> 17.650 -1315.0
## + b18a 1 0.04480 17.605 -1314.1
## + d1a 1 0.04281 17.607 -1314.0
## + b16a 1 0.03148 17.619 -1313.8
## + h1 1 0.01659 17.633 -1313.4
## + c1 1 0.00364 17.646 -1313.1
## - smostt 2 0.25653 17.907 -1313.0
## + ter_in 2 0.06317 17.587 -1312.5
## + age_group 4 0.18533 17.465 -1311.5
## + policy_a 2 0.00942 17.641 -1311.2
## - selfhealth 1 0.45559 18.106 -1306.3
## - educ2 2 0.97460 18.625 -1296.5##
## Call:
## lm(formula = logC ~ educ2 + selfhealth + smostt, data = t2)
##
## Coefficients:
## (Intercept) educ22 educ23 selfhealthgood
## 4.66055 0.04186 0.13784 0.06804
## smosttmedium smosttheavy
## 0.04924 0.06749Mo hinh va bieu do
library(BMA)## Warning: package 'BMA' was built under R version 3.2.5## Loading required package: survival## Warning: package 'survival' was built under R version 3.2.5## Loading required package: leaps## Warning: package 'leaps' was built under R version 3.2.5## Loading required package: robustbase## Warning: package 'robustbase' was built under R version 3.2.5##
## Attaching package: 'robustbase'## The following object is masked from 'package:survival':
##
## heart## Loading required package: inline## Warning: package 'inline' was built under R version 3.2.5## Loading required package: rrcov## Warning: package 'rrcov' was built under R version 3.2.5## Scalable Robust Estimators with High Breakdown Point (version 1.4-3)bma = bicreg(xvars, yvar, strict=FALSE, OR=20)
summary(bma)##
## Call:
## bicreg(x = xvars, y = yvar, strict = FALSE, OR = 20)
##
##
## 18 models were selected
## Best 5 models (cumulative posterior probability = 0.6577 ):
##
## p!=0 EV SD model 1 model 2
## Intercept 100.0 4.7382241 0.019749 4.73789 4.74997
## age_groupgr3039 3.2 0.0009045 0.007020 . .
## age_groupgr4049 2.3 -0.0003877 0.004621 . .
## age_groupgr5059 11.8 -0.0062094 0.019644 . -0.05237
## age_group60plus 0.0 0.0000000 0.000000 . .
## d1amarried 2.9 -0.0005841 0.005026 . .
## educ22 5.5 0.0021859 0.011006 . .
## educ23 100.0 0.0890641 0.024483 0.08723 0.07691
## ter_in2 5.2 0.0016958 0.008939 . .
## ter_in3 1.9 -0.0001157 0.003218 . .
## selfhealthgood 97.1 0.0669088 0.023414 0.06915 0.06750
## h12 2.4 -0.0003771 0.004056 . .
## b16a1 2.2 -0.0003839 0.004998 . .
## b18abad 3.0 0.0006232 0.005127 . .
## c12 9.3 -0.0047015 0.017411 . .
## smosttmedium 2.2 0.0002573 0.003529 . .
## smosttheavy 5.5 0.0017750 0.009018 . .
## policy_a1 1.9 -0.0001179 0.003229 . .
## policy_a2 2.2 -0.0003480 0.004744 . .
##
## nVar 2 3
## r2 0.068 0.075
## BIC -17.48522 -14.77429
## post prob 0.378 0.097
## model 3 model 4 model 5
## Intercept 4.74056 4.71247 4.72378
## age_groupgr3039 . . .
## age_groupgr4049 . . .
## age_groupgr5059 . . .
## age_group60plus . . .
## d1amarried . . .
## educ22 . 0.03949 .
## educ23 0.09695 0.11551 0.09530
## ter_in2 . . .
## ter_in3 . . .
## selfhealthgood 0.07226 0.06399 0.07070
## h12 . . .
## b16a1 . . .
## b18abad . . .
## c12 -0.05038 . .
## smosttmedium . . .
## smosttheavy . . 0.03250
## policy_a1 . . .
## policy_a2 . . .
##
## nVar 3 3 3
## r2 0.074 0.073 0.073
## BIC -14.17195 -13.64279 -13.61567
## post prob 0.072 0.055 0.055image.plot(bma)
install.packages(“relaimpo”)
which vars is importants
model1 = lm(logC~., data=t2)
library(relaimpo)## Warning: package 'relaimpo' was built under R version 3.2.5## Loading required package: MASS## Loading required package: boot##
## Attaching package: 'boot'## The following object is masked from 'package:robustbase':
##
## salinity## The following object is masked from 'package:survival':
##
## aml## Loading required package: survey## Warning: package 'survey' was built under R version 3.2.5## Loading required package: grid## Loading required package: Matrix##
## Attaching package: 'survey'## The following object is masked from 'package:graphics':
##
## dotchart## Loading required package: mitools## Warning: package 'mitools' was built under R version 3.2.5## This is the global version of package relaimpo.## If you are a non-US user, a version with the interesting additional metric pmvd is available## from Ulrike Groempings web site at prof.beuth-hochschule.de/groemping.metrics=calc.relimp(model1, type=c("lmg"))
metrics## Response variable: logC
## Total response variance: 0.04621148
## Analysis based on 419 observations
##
## 18 Regressors:
## Some regressors combined in groups:
## Group age_group : age_groupgr3039 age_groupgr4049 age_groupgr5059 age_group60plus
## Group educ2 : educ22 educ23
## Group ter_in : ter_in2 ter_in3
## Group smostt : smosttmedium smosttheavy
## Group policy_a : policy_a1 policy_a2
##
## Relative importance of 11 (groups of) regressors assessed:
## age_group educ2 ter_in smostt policy_a d1a selfhealth h1 b16a b18a c1
##
## Proportion of variance explained by model: 10.44%
## Metrics are not normalized (rela=FALSE).
##
## Relative importance metrics:
##
## lmg
## age_group 0.0200875810
## educ2 0.0373578363
## ter_in 0.0031377768
## smostt 0.0052965668
## policy_a 0.0003172111
## d1a 0.0048398829
## selfhealth 0.0254845987
## h1 0.0016147509
## b16a 0.0018942797
## b18a 0.0032299955
## c1 0.0011670414
##
## Average coefficients for different model sizes:
##
## 1group 2groups 3groups 4groups
## age_groupgr3039 0.000446899 0.0008963928 0.0013424084 0.001808269
## age_groupgr4049 -0.064030447 -0.0603790335 -0.0566866696 -0.052960998
## age_groupgr5059 -0.105138831 -0.1001093303 -0.0950382670 -0.089944017
## age_group60plus -0.085031354 -0.0783814902 -0.0717668141 -0.065209248
## d1a -0.049692105 -0.0431291919 -0.0372362694 -0.031934413
## educ22 0.053227138 0.0503693935 0.0475680649 0.044839655
## educ23 0.134294102 0.1316710282 0.1290289035 0.126395842
## ter_in2 0.035622950 0.0342790552 0.0325963851 0.030745648
## ter_in3 0.016707055 0.0156455026 0.0141336859 0.012333854
## selfhealth 0.079762494 0.0769257553 0.0745093447 0.072451722
## h1 -0.021083514 -0.0196572100 -0.0185791047 -0.017787456
## b16a -0.019892078 -0.0218696173 -0.0236920850 -0.025331012
## b18a 0.023588910 0.0247031036 0.0255310866 0.026130881
## c1 -0.003261323 -0.0107447997 -0.0160511808 -0.019482341
## smosttmedium 0.007781878 0.0143748671 0.0198925641 0.024521555
## smosttheavy 0.008369407 0.0188615609 0.0276980684 0.035129151
## policy_a1 0.001323079 0.0002564719 -0.0002893969 -0.000502695
## policy_a2 -0.011580535 -0.0095336537 -0.0078323045 -0.006491841
## 5groups 6groups 7groups 8groups
## age_groupgr3039 0.0023082728 0.0028499769 0.003435992 0.0040652349
## age_groupgr4049 -0.0492204580 -0.0454919243 -0.041808878 -0.0382101617
## age_groupgr5059 -0.0848499046 -0.0797830184 -0.074773556 -0.0698546803
## age_group60plus -0.0587331396 -0.0523649773 -0.046133574 -0.0400706575
## d1a -0.0271558212 -0.0228440585 -0.018953664 -0.0154489379
## educ22 0.0421981705 0.0396547208 0.037217175 0.0348898978
## educ23 0.1237989249 0.1212629615 0.118809501 0.1164560391
## ter_in2 0.0288544934 0.0270152733 0.025292217 0.0237279393
## ter_in3 0.0103677039 0.0083247489 0.006269863 0.0042497443
## selfhealth 0.0706970549 0.0691951011 0.067900557 0.0667719462
## h1 -0.0172302922 -0.0168642435 -0.016653073 -0.0165660642
## b16a -0.0267628582 -0.0279691080 -0.028936452 -0.0296567771
## b18a 0.0265424083 0.0267913073 0.026892832 0.0268552808
## c1 -0.0213134307 -0.0217920759 -0.021137716 -0.0195409002
## smosttmedium 0.0284254962 0.0317447902 0.034597387 0.0370803803
## smosttheavy 0.0413823166 0.0466607936 0.051143311 0.0549848741
## policy_a1 -0.0005310473 -0.0004875677 -0.000456894 -0.0005007053
## policy_a2 -0.0055153556 -0.0048952793 -0.004614693 -0.0046481014
## 9groups 10groups 11groups
## age_groupgr3039 0.0047336777 0.0054347439 0.006159504
## age_groupgr4049 -0.0347392171 -0.0314435986 -0.028374533
## age_groupgr5059 -0.0650627606 -0.0604378063 -0.056023850
## age_group60plus -0.0342117233 -0.0285969329 -0.023271812
## d1a -0.0123018161 -0.0094889176 -0.006987929
## educ22 0.0326735764 0.0305651526 0.028557817
## educ23 0.1142153932 0.1120951688 0.110097250
## ter_in2 0.0223491894 0.0211716777 0.020203773
## ter_in3 0.0022980655 0.0004391681 -0.001309746
## selfhealth 0.0657700926 0.0648562173 0.063989677
## h1 -0.0165764354 -0.0166599949 -0.016794248
## b16a -0.0301268306 -0.0303475763 -0.030323422
## b18a 0.0266826486 0.0263763904 0.025936263
## c1 -0.0171625509 -0.0141333718 -0.010553656
## smosttmedium 0.0392722063 0.0412353582 0.043019675
## smosttheavy 0.0583183172 0.0612565081 0.063895226
## policy_a1 -0.0006625225 -0.0009717927 -0.001447362
## policy_a2 -0.0049617275 -0.0055135566 -0.006253434#bootstrap
boot = boot.relimp(model1, b=1000,
type=c("lmg"), fixed=F)
booteval.relimp(boot, typesel=c("lmg"),
level=0.9, bty="perc", nodiff=T)## Response variable: logC
## Total response variance: 0.04621148
## Analysis based on 419 observations
##
## 18 Regressors:
## Some regressors combined in groups:
## Group age_group : age_groupgr3039 age_groupgr4049 age_groupgr5059 age_group60plus
## Group educ2 : educ22 educ23
## Group ter_in : ter_in2 ter_in3
## Group smostt : smosttmedium smosttheavy
## Group policy_a : policy_a1 policy_a2
##
## Relative importance of 11 (groups of) regressors assessed:
## age_group educ2 ter_in smostt policy_a d1a selfhealth h1 b16a b18a c1
##
## Proportion of variance explained by model: 10.44%
## Metrics are not normalized (rela=FALSE).
##
## Relative importance metrics:
##
## lmg
## age_group 0.0200875810
## educ2 0.0373578363
## ter_in 0.0031377768
## smostt 0.0052965668
## policy_a 0.0003172111
## d1a 0.0048398829
## selfhealth 0.0254845987
## h1 0.0016147509
## b16a 0.0018942797
## b18a 0.0032299955
## c1 0.0011670414
##
## Average coefficients for different model sizes:
##
## 1group 2groups 3groups 4groups
## age_groupgr3039 0.000446899 0.0008963928 0.0013424084 0.001808269
## age_groupgr4049 -0.064030447 -0.0603790335 -0.0566866696 -0.052960998
## age_groupgr5059 -0.105138831 -0.1001093303 -0.0950382670 -0.089944017
## age_group60plus -0.085031354 -0.0783814902 -0.0717668141 -0.065209248
## d1a -0.049692105 -0.0431291919 -0.0372362694 -0.031934413
## educ22 0.053227138 0.0503693935 0.0475680649 0.044839655
## educ23 0.134294102 0.1316710282 0.1290289035 0.126395842
## ter_in2 0.035622950 0.0342790552 0.0325963851 0.030745648
## ter_in3 0.016707055 0.0156455026 0.0141336859 0.012333854
## selfhealth 0.079762494 0.0769257553 0.0745093447 0.072451722
## h1 -0.021083514 -0.0196572100 -0.0185791047 -0.017787456
## b16a -0.019892078 -0.0218696173 -0.0236920850 -0.025331012
## b18a 0.023588910 0.0247031036 0.0255310866 0.026130881
## c1 -0.003261323 -0.0107447997 -0.0160511808 -0.019482341
## smosttmedium 0.007781878 0.0143748671 0.0198925641 0.024521555
## smosttheavy 0.008369407 0.0188615609 0.0276980684 0.035129151
## policy_a1 0.001323079 0.0002564719 -0.0002893969 -0.000502695
## policy_a2 -0.011580535 -0.0095336537 -0.0078323045 -0.006491841
## 5groups 6groups 7groups 8groups
## age_groupgr3039 0.0023082728 0.0028499769 0.003435992 0.0040652349
## age_groupgr4049 -0.0492204580 -0.0454919243 -0.041808878 -0.0382101617
## age_groupgr5059 -0.0848499046 -0.0797830184 -0.074773556 -0.0698546803
## age_group60plus -0.0587331396 -0.0523649773 -0.046133574 -0.0400706575
## d1a -0.0271558212 -0.0228440585 -0.018953664 -0.0154489379
## educ22 0.0421981705 0.0396547208 0.037217175 0.0348898978
## educ23 0.1237989249 0.1212629615 0.118809501 0.1164560391
## ter_in2 0.0288544934 0.0270152733 0.025292217 0.0237279393
## ter_in3 0.0103677039 0.0083247489 0.006269863 0.0042497443
## selfhealth 0.0706970549 0.0691951011 0.067900557 0.0667719462
## h1 -0.0172302922 -0.0168642435 -0.016653073 -0.0165660642
## b16a -0.0267628582 -0.0279691080 -0.028936452 -0.0296567771
## b18a 0.0265424083 0.0267913073 0.026892832 0.0268552808
## c1 -0.0213134307 -0.0217920759 -0.021137716 -0.0195409002
## smosttmedium 0.0284254962 0.0317447902 0.034597387 0.0370803803
## smosttheavy 0.0413823166 0.0466607936 0.051143311 0.0549848741
## policy_a1 -0.0005310473 -0.0004875677 -0.000456894 -0.0005007053
## policy_a2 -0.0055153556 -0.0048952793 -0.004614693 -0.0046481014
## 9groups 10groups 11groups
## age_groupgr3039 0.0047336777 0.0054347439 0.006159504
## age_groupgr4049 -0.0347392171 -0.0314435986 -0.028374533
## age_groupgr5059 -0.0650627606 -0.0604378063 -0.056023850
## age_group60plus -0.0342117233 -0.0285969329 -0.023271812
## d1a -0.0123018161 -0.0094889176 -0.006987929
## educ22 0.0326735764 0.0305651526 0.028557817
## educ23 0.1142153932 0.1120951688 0.110097250
## ter_in2 0.0223491894 0.0211716777 0.020203773
## ter_in3 0.0022980655 0.0004391681 -0.001309746
## selfhealth 0.0657700926 0.0648562173 0.063989677
## h1 -0.0165764354 -0.0166599949 -0.016794248
## b16a -0.0301268306 -0.0303475763 -0.030323422
## b18a 0.0266826486 0.0263763904 0.025936263
## c1 -0.0171625509 -0.0141333718 -0.010553656
## smosttmedium 0.0392722063 0.0412353582 0.043019675
## smosttheavy 0.0583183172 0.0612565081 0.063895226
## policy_a1 -0.0006625225 -0.0009717927 -0.001447362
## policy_a2 -0.0049617275 -0.0055135566 -0.006253434
##
##
## Confidence interval information ( 1000 bootstrap replicates, bty= perc ):
## Relative Contributions with confidence intervals:
##
## Lower Upper
## percentage 0.9 0.9 0.9
## age_group.lmg 0.0201 ABCDE______ 0.0102 0.0515
## educ2.lmg 0.0374 ABC________ 0.0171 0.0689
## ter_in.lmg 0.0031 __CDEFGHIJ_ 0.0011 0.0212
## smostt.lmg 0.0053 __CDEFGHI__ 0.0021 0.0229
## policy_a.lmg 0.0003 ___DEFGHIJK 0.0006 0.0126
## d1a.lmg 0.0048 ___DEFGHIJ_ 0.0013 0.0166
## selfhealth.lmg 0.0255 ABCDE______ 0.0080 0.0540
## h1.lmg 0.0016 ___DEFGHIJK 0.0002 0.0133
## b16a.lmg 0.0019 ___DEFGHIJK 0.0002 0.0146
## b18a.lmg 0.0032 __CDEFGHIJK 0.0002 0.0173
## c1.lmg 0.0012 ____EFGHIJK 0.0006 0.0099
##
## Letters indicate the ranks covered by bootstrap CIs.
## (Rank bootstrap confidence intervals always obtained by percentile method)
## CAUTION: Bootstrap confidence intervals can be somewhat liberal.reg = lm(logC ~ c1+ter_in+age_group+educ2+selfhealth+smostt+policy_a, data=t2)
summary(reg)##
## Call:
## lm(formula = logC ~ c1 + ter_in + age_group + educ2 + selfhealth +
## smostt + policy_a, data = t2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.55776 -0.14711 -0.00187 0.14897 0.52901
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.697343 0.043838 107.153 < 2e-16 ***
## c12 -0.008962 0.040115 -0.223 0.82332
## ter_in2 0.020518 0.025860 0.793 0.42800
## ter_in3 -0.004564 0.026111 -0.175 0.86132
## age_groupgr3039 0.002269 0.031147 0.073 0.94196
## age_groupgr4049 -0.033327 0.029619 -1.125 0.26118
## age_groupgr5059 -0.057855 0.033655 -1.719 0.08637 .
## age_group60plus -0.026090 0.048357 -0.540 0.58982
## educ22 0.026184 0.028225 0.928 0.35413
## educ23 0.112617 0.034579 3.257 0.00122 **
## selfhealthgood 0.064255 0.021373 3.006 0.00281 **
## smosttmedium 0.043157 0.033282 1.297 0.19547
## smosttheavy 0.064494 0.034563 1.866 0.06277 .
## policy_a1 -0.005284 0.023712 -0.223 0.82376
## policy_a2 -0.005389 0.030000 -0.180 0.85753
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2076 on 404 degrees of freedom
## Multiple R-squared: 0.09842, Adjusted R-squared: 0.06718
## F-statistic: 3.15 on 14 and 404 DF, p-value: 0.000104anova(reg)## Analysis of Variance Table
##
## Response: logC
## Df Sum Sq Mean Sq F value Pr(>F)
## c1 1 0.0005 0.00053 0.0123 0.911640
## ter_in 2 0.0895 0.04473 1.0377 0.355224
## age_group 4 0.7226 0.18066 4.1909 0.002454 **
## educ2 2 0.5551 0.27755 6.4386 0.001768 **
## selfhealth 1 0.3794 0.37937 8.8006 0.003190 **
## smostt 2 0.1513 0.07564 1.7547 0.174282
## policy_a 2 0.0028 0.00139 0.0323 0.968251
## Residuals 404 17.4153 0.04311
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1par(mfrow=c(2,2))plot(reg)
install.packages(“lsmeans”)
library(lsmeans)## Warning: package 'lsmeans' was built under R version 3.2.5## Loading required package: estimability## Warning: package 'estimability' was built under R version 3.2.5lsmeans(reg, "age_group")## age_group lsmean SE df lower.CL upper.CL
## group18-29 4.808900 0.02271738 404 4.764241 4.853559
## gr3039 4.811170 0.03060186 404 4.751011 4.871328
## gr4049 4.775574 0.02755691 404 4.721401 4.829746
## gr5059 4.751045 0.03094551 404 4.690211 4.811880
## 60plus 4.782811 0.04712900 404 4.690162 4.875460
##
## Results are averaged over the levels of: c1, ter_in, educ2, selfhealth, smostt, policy_a
## Confidence level used: 0.95educcation
lsmeans(reg, "educ2")## educ2 lsmean SE df lower.CL upper.CL
## 1 4.739633 0.02856822 404 4.683472 4.795794
## 2 4.765817 0.02267017 404 4.721251 4.810383
## 3 4.852250 0.02749980 404 4.798189 4.906310
##
## Results are averaged over the levels of: c1, ter_in, age_group, selfhealth, smostt, policy_a
## Confidence level used: 0.95Selfhealth
lsmeans(reg, "selfhealth")## selfhealth lsmean SE df lower.CL upper.CL
## notwell 4.753772 0.02209456 404 4.710338 4.797207
## good 4.818028 0.02310997 404 4.772597 4.863458
##
## Results are averaged over the levels of: c1, ter_in, age_group, educ2, smostt, policy_a
## Confidence level used: 0.95Anti cam
lsmeans(reg, "policy_a")## policy_a lsmean SE df lower.CL upper.CL
## 0 4.789458 0.02251976 404 4.745187 4.833728
## 1 4.784173 0.02408767 404 4.736821 4.831526
## 2 4.784069 0.03074085 404 4.723637 4.844501
##
## Results are averaged over the levels of: c1, ter_in, age_group, educ2, selfhealth, smostt
## Confidence level used: 0.95prequency
lsmeans(reg, "smostt")## smostt lsmean SE df lower.CL upper.CL
## light 4.750016 0.02338176 404 4.704051 4.795981
## medium 4.793174 0.02816266 404 4.737810 4.848537
## heavy 4.814510 0.02852786 404 4.758429 4.870592
##
## Results are averaged over the levels of: c1, ter_in, age_group, educ2, selfhealth, policy_a
## Confidence level used: 0.95- Binh Thang Ref: -Coppied: Prof Nguyen Van Tuan, (tuanvannguyen.blogspot.kr)