【原创】R语言逻辑回归NRI、IDI 指标比较案例 附代码数据.docx
《【原创】R语言逻辑回归NRI、IDI 指标比较案例 附代码数据.docx》由会员分享,可在线阅读,更多相关《【原创】R语言逻辑回归NRI、IDI 指标比较案例 附代码数据.docx(7页珍藏版)》请在冰点文库上搜索。
【原创】附代码数据
有问题到淘宝找“大数据部落”就可以了
逻辑回归NRI、IDI指标比较案例
head(data)
##是否复发术前PSACAPRA-S评分725.2261014.161328.55
##1
0
3
2
2
2
1
##2
0
3
2
2
3
1
##3
0
1
1
1
3
1
##4
0
2
2
2
1
2
##5
0
2
2
1
1
2
##6
0
2
2
1
2
1
三个logistic回归模型与只有PSA+CAPRAscore基础模型
加入了725,1014和1328,分别建立了三个logistic回归模型,分别比较NRI、IDI指标。
原始模型
logitMod<-glm(是否复发~术前PSA+`CAPRA-S评分`,data=data,family=binomial(link="logit"))summary(logitMod)
DevianceResiduals:
Min 1Q Median 3Q Max
-1.8076-1.0169-0.4592 0.7041 2.1457
Coefficients:
EstimateStd.ErrorzvaluePr(>|z|)
(Intercept) -4.0032 0.9489-4.2192.45e-05***
术前PSA 0.1481 0.3991 0.3710.710537
`CAPRA-S评分` 1.6585 0.4641 3.5740.000352***
---
Signif.codes:
0'***'0.001'**'0.01'*'0.05'.'0.1''1
(Dispersionparameterforbinomialfamilytakentobe1)
Nulldeviance:
131.85on98degreesoffreedomResidualdeviance:
106.78on96degreesoffreedomAIC:
112.78
NumberofFisherScoringiterations:
4
725.226
logitMod1<-glm(是否复发~术前PSA+`CAPRA-S评分`+`725.226`,data=data,family=binomial(link="logit"))summary(logitMod1)
DevianceResiduals:
Min 1Q Median 3Q Max
-1.7219-0.8988 -0.4221 0.7178 2.8554
Coefficients:
EstimateStd.ErrorzvaluePr(>|z|)
(Intercept) -7.1845 1.4842 -4.8411.29e-06***
术前PSA -0.2366 0.4501 -0.5260.599055
`CAPRA-S评分` 1.9122 0.5452 3.5070.000453***
`725.226` 1.6861 0.4451 3.7880.000152***
---
Signif.codes:
0'***'0.001'**'0.01'*'0.05'.'0.1''1
(Dispersionparameterforbinomialfamilytakentobe1)
Nulldeviance:
131.851on98degreesoffreedomResidualdeviance:
88.256on95degreesoffreedomAIC:
96.256
NumberofFisherScoringiterations:
5
1014.16
#1014.16
logitMod2<-glm(是否复发~术前PSA+`CAPRA-S评分`+`1014.16`,data=data,family=binomial(link="logit"))summary(logitMod2)
DevianceResiduals:
Min 1Q Median 3Q Max
-1.7453-0.5387 -0.4765 0.7288 2.6769
Coefficients:
EstimateStd.ErrorzvaluePr(>|z|)
(Intercept) -0.9433 1.2427 -0.7590.447792
术前PSA 0.0903 0.4575 0.1970.843530
`CAPRA-S评分` 1.6073 0.5108 3.1470.001650**
`1014.16` -1.4363 0.4010-3.5820.000341***
---
Signif.codes:
0'***'0.001'**'0.01'*'0.05'.'0.1''1
(Dispersionparameterforbinomialfamilytakentobe1)
Nulldeviance:
131.851on98degreesoffreedomResidualdeviance:
90.939on95degreesoffreedomAIC:
98.939
NumberofFisherScoringiterations:
5
1328.55
#logitMod3<-glm(是否复发~术前PSA+`CAPRA-S评分`+`1328.55`,data=data,family=binomial(link="logit"))
summary(logitMod3)
DevianceResiduals:
Min 1Q Median 3Q Max
-1.7580-0.7114 -0.4417 0.6742 2.7821
Coefficients:
EstimateStd.ErrorzvaluePr(>|z|)
(Intercept) -7.4556 1.4946 -4.9886.09e-07***
术前PSA
0.1220
0.4851
0.2510.801475
`CAPRA-S评分`
1.5707
0.5431
2.8920.003824**
`1328.55`
1.7919
0.4662
3.8440.000121***
---
Signif.codes:
0'***'0.001'**'0.01'*'0.05'.'0.1''1
(Dispersionparameterforbinomialfamilytakentobe1)
Nulldeviance:
131.851on98degreesoffreedomResidualdeviance:
86.712on95degreesoffreedomAIC:
94.712
NumberofFisherScoringiterations:
5
与725.226模型比较的NRI、IDI
Reclassificationtable
Outcome:
absent
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
57
1
2
[0.5,1]
0
3
0
Outcome:
present
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
13
10
43
[0.5,1]
3
12
20
CombinedData
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
70
11
14
[0.5,1]
3
15
17
NRI(Categorical)[95%CI]:
0.1678[-0.0115-0.3472];p-value:
0.06667
NRI(Continuous)[95%CI]:
0.9422[0.5822-1.3022];p-value:
0
IDI[95%CI]:
0.1584[0.0812-0.2356];p-value:
6e-05
1014.16的NRI、IDI
Reclassificationtable
Outcome:
absent
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
56
2
3
[0.5,1]
1
2
33
Outcome:
present
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
12
11
48
[0.5,1]
2
13
13
CombinedData
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
68
13
16
[0.5,1]
3
15
17
NRI(Categorical)[95%CI]:
0.2204[0.0416-0.3993];p-value:
0.01571
NRI(Continuous)[95%CI]:
0.4107[0.0161-0.8053];p-value:
0.04135
IDI[95%CI]:
0.1459[0.0713-0.2206];p-value:
0.00013
1328.55的NRI、IDI
Reclassificationtable
Outcome:
absent
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
57
1
2
[0.5,1]
1
2
33
Outcome:
present
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
14
9
39
[0.5,1]
1
14
7
CombinedData
UpdatedModel
InitialModel[0,0.5)[0.5,1]%reclassified
[0,0.5)
71
10
12
[0.5,1]
2
16
11
NRI(Categorical)[95%CI]:
0.2105[0.055-0.3661];p-value:
0.00797
NRI(Continuous)[95%CI]:
0.679[0.3019-1.0562];p-value:
0.00042
IDI[95%CI]:
0.1753[0.0969-0.2538];p-value:
1e-05