实例介绍
【实例简介】
对于理解多重填补法有很大帮助，特别是对初学者有很大帮助。
20036 2 79· PMM u∑ (reg ression method 22 PS (Xm,中1),(Xm,中2),K P(Xm,中X gIs logit(pi)=B+β1X1+B2X2+…+ t B-1X P(Xm|X,中 (4)pi=Pr(R=0X1,X2, P(中Xn,X们+1 git(p g p 3 logit(pi) m> (3 Bootstrap( approx imat e Bayesian Bootstrap, AB B) 0 ele)2….m n 1 MI n 1 e=-∑6 23 MCMC M CM C (Imputation step) (10 P(XmXo,中 1∑(t-6 μ1,P2 n nLs B l1 (5 6)102…fn 2 OB ∑12Xob、X OB bs =1 21=12 11(x1-1) 95% 6土2 12 12 (7 o(254-2012(osterior steric journal el e± 13 ctronic pu iblishing House. All rights res http://www.cnki.net 80· 20036 (13) m Ow 1(m-1)1+ (m+1) (14) MI 8 r+2/y+3 (15 r MI 15 M CM C MI MI (v=∞) l; 10 MI 表1多重填补的相对效率 10% 30% 509 70 0.96770.93750.90910.85710.8108 0.98040.96150.94340.90910.8772 00. 0.98040.97090.95240.9346 20 0.99500.99010.98520.97560.9662 MI 2.13 MI list wise dele M tion, single imput at ion) MI MI MI I Barnard J, Meng X L. A pplicat ions of mult iple im put at ion in medi cal study: from ADiS to NHANESI J]. Stat istical Met hods in Medt MI ex pect ation maximizat ion, al research,1999,8(1):17~36 2 Barnard J, Rubin D B. Small sample degrees of freedom with multt EM) ole im put at ion[ J]. Biomet rika, 1999, 86(4): 948-955 MI 3 Fienberg SE, Willenborg L C. Introduct ion to the sp C1994-2012 China Academic Journal Electronic Publishing flow imit at ion. me thods for protecting the confident iality,of st disclosure 20036 2 81· tistical dat a[ J Journal of O fficial St at ist ics, 1998, 14(4): 337- 9 Wangn, Robins J M. Large- sample theory for parametric multi- ole imput at ion procedures[ J]. Biometrika, 1998, 85(4): 935-948 4 Horton N J, Laird N M. M ax imum lik elihoo d analy sis of generalized 10 James H, Anne J, Gary K, et al. Amelia: A Program for M issing linear models with missing covariates J ]. Statist ical Methods in Data[S]. Department of Govemm ent, H arvard U niversity, 1999 Medical Research, 1999, 8(1): 37-50 5 King G, Honaker J, Joseph A, et al. Analy zing incomplete polit ical 11 Aidan M. M ult iple imput at ion for missing data using the" Solas for science data: an alt ernat ive algorithm for multiple imputation[ JI the m issing da ta a nalysis"sof tw are applicat ion[S]. Conference of A merican Polit ical Scien ce Review, 2001, 95( 1): 46-69 Europe an S tatisticians. 1999. 2-8 6 M X L, Rubin D B. Performing likelihood rat ion tests w ith mut 12 Nicholas J H, Stuart R L. Multiple imput at ion in Pract ices: co- tiple imput ed da ta sets[ J]. Biometrika, 1992, 79(1 ): 103-111 parison of softw are packages for the re gression models w it h missing 7 Robins JM, Wang n. Inference for imput at ion est im at ors[ JI variables[ J]. The Am erican Sta tist ician, 2001, 55(3): 244-254 Biomet rika,2000,87(1):113~124 13 Y ang CY. Multiple imputa tion for missing dat a: concepts and new 8 Rubin D B. Muliple imputa tion: a primer[ J]. St at is icaI M et hods development[S]. SAS Inst it ute Inc. 1999.4-9 in Medical Research, 1999, 8(1):3-15 20020912) 20种住院死亡疾病年龄统计 赵炜!胡军!马虹万晓阳 】死亡年龄统计 :R195.4 B 1006-5253(2003)02008H01 1971-1998 43 20 3176 66.1%,2034 l142 20 20 1971-1998 4803 表120种疾病死亡病例年龄统计 /% 0 11 31 41 61 71 178 0.6 8.4 24.2 14.6 18.0 16.9 14.6 15.6 27.3 14. 14.1 13.3 7.0 2.7 6.3 16.2 13.5 15.3 15.3 17. 17.8 29.2 23.9 14.6 2.3 2.3 10.5 15 23.3 23.3 15 4.3 14.5 20.3 23.2 5.4 4 453 4.0 9.5 28.9 34.7 18.3 0.4 20.2 33.5 24.6 5 0.6 0.6 4.5 18.0 25.0 30.0 l11 2.7 6.3 16.2 13.5 15.3 15.3 4.5 9.7 9.4 2.1 12.0 27.5 20. 6.1 303 0.3 5.6 8.9 27.7 33.7 17.5 296 2.0 2.4 8.1 21.6 32.8 24.3 8.8 144 0.7 6.3 28.5 31.9 22.9 9.0 25.3 37.4 28.6 6.6 4 9.7 6.9 6.9 25.0 30.6 13.9 4.7 18.6 30.2 23.3 16.3 3.2 16.8 25.3 27.4 26.3 1.2 4.9 2.2 25.6 30.5 22.0 100.0 1110003 202 C1994-20f2ChinaAcademicJournaleLectronicPublishingHouse.Allrightsreservedhttp://www.cnti.net 【实例截图】
【核心代码】
对于理解多重填补法有很大帮助，特别是对初学者有很大帮助。
20036 2 79· PMM u∑ (reg ression method 22 PS (Xm,中1),(Xm,中2),K P(Xm,中X gIs logit(pi)=B+β1X1+B2X2+…+ t B-1X P(Xm|X,中 (4)pi=Pr(R=0X1,X2, P(中Xn,X们+1 git(p g p 3 logit(pi) m> (3 Bootstrap( approx imat e Bayesian Bootstrap, AB B) 0 ele)2….m n 1 MI n 1 e=-∑6 23 MCMC M CM C (Imputation step) (10 P(XmXo,中 1∑(t-6 μ1,P2 n nLs B l1 (5 6)102…fn 2 OB ∑12Xob、X OB bs =1 21=12 11(x1-1) 95% 6土2 12 12 (7 o(254-2012(osterior steric journal el e± 13 ctronic pu iblishing House. All rights res http://www.cnki.net 80· 20036 (13) m Ow 1(m-1)1+ (m+1) (14) MI 8 r+2/y+3 (15 r MI 15 M CM C MI MI (v=∞) l; 10 MI 表1多重填补的相对效率 10% 30% 509 70 0.96770.93750.90910.85710.8108 0.98040.96150.94340.90910.8772 00. 0.98040.97090.95240.9346 20 0.99500.99010.98520.97560.9662 MI 2.13 MI list wise dele M tion, single imput at ion) MI MI MI I Barnard J, Meng X L. A pplicat ions of mult iple im put at ion in medi cal study: from ADiS to NHANESI J]. Stat istical Met hods in Medt MI ex pect ation maximizat ion, al research,1999,8(1):17~36 2 Barnard J, Rubin D B. Small sample degrees of freedom with multt EM) ole im put at ion[ J]. Biomet rika, 1999, 86(4): 948-955 MI 3 Fienberg SE, Willenborg L C. Introduct ion to the sp C1994-2012 China Academic Journal Electronic Publishing flow imit at ion. me thods for protecting the confident iality,of st disclosure 20036 2 81· tistical dat a[ J Journal of O fficial St at ist ics, 1998, 14(4): 337- 9 Wangn, Robins J M. Large- sample theory for parametric multi- ole imput at ion procedures[ J]. Biometrika, 1998, 85(4): 935-948 4 Horton N J, Laird N M. M ax imum lik elihoo d analy sis of generalized 10 James H, Anne J, Gary K, et al. Amelia: A Program for M issing linear models with missing covariates J ]. Statist ical Methods in Data[S]. Department of Govemm ent, H arvard U niversity, 1999 Medical Research, 1999, 8(1): 37-50 5 King G, Honaker J, Joseph A, et al. Analy zing incomplete polit ical 11 Aidan M. M ult iple imput at ion for missing data using the" Solas for science data: an alt ernat ive algorithm for multiple imputation[ JI the m issing da ta a nalysis"sof tw are applicat ion[S]. Conference of A merican Polit ical Scien ce Review, 2001, 95( 1): 46-69 Europe an S tatisticians. 1999. 2-8 6 M X L, Rubin D B. Performing likelihood rat ion tests w ith mut 12 Nicholas J H, Stuart R L. Multiple imput at ion in Pract ices: co- tiple imput ed da ta sets[ J]. Biometrika, 1992, 79(1 ): 103-111 parison of softw are packages for the re gression models w it h missing 7 Robins JM, Wang n. Inference for imput at ion est im at ors[ JI variables[ J]. The Am erican Sta tist ician, 2001, 55(3): 244-254 Biomet rika,2000,87(1):113~124 13 Y ang CY. Multiple imputa tion for missing dat a: concepts and new 8 Rubin D B. Muliple imputa tion: a primer[ J]. St at is icaI M et hods development[S]. SAS Inst it ute Inc. 1999.4-9 in Medical Research, 1999, 8(1):3-15 20020912) 20种住院死亡疾病年龄统计 赵炜!胡军!马虹万晓阳 】死亡年龄统计 :R195.4 B 1006-5253(2003)02008H01 1971-1998 43 20 3176 66.1%,2034 l142 20 20 1971-1998 4803 表120种疾病死亡病例年龄统计 /% 0 11 31 41 61 71 178 0.6 8.4 24.2 14.6 18.0 16.9 14.6 15.6 27.3 14. 14.1 13.3 7.0 2.7 6.3 16.2 13.5 15.3 15.3 17. 17.8 29.2 23.9 14.6 2.3 2.3 10.5 15 23.3 23.3 15 4.3 14.5 20.3 23.2 5.4 4 453 4.0 9.5 28.9 34.7 18.3 0.4 20.2 33.5 24.6 5 0.6 0.6 4.5 18.0 25.0 30.0 l11 2.7 6.3 16.2 13.5 15.3 15.3 4.5 9.7 9.4 2.1 12.0 27.5 20. 6.1 303 0.3 5.6 8.9 27.7 33.7 17.5 296 2.0 2.4 8.1 21.6 32.8 24.3 8.8 144 0.7 6.3 28.5 31.9 22.9 9.0 25.3 37.4 28.6 6.6 4 9.7 6.9 6.9 25.0 30.6 13.9 4.7 18.6 30.2 23.3 16.3 3.2 16.8 25.3 27.4 26.3 1.2 4.9 2.2 25.6 30.5 22.0 100.0 1110003 202 C1994-20f2ChinaAcademicJournaleLectronicPublishingHouse.Allrightsreservedhttp://www.cnti.net 【实例截图】
【核心代码】
标签：
好例子网口号：伸出你的我的手 — 分享！
小贴士
感谢您为本站写下的评论，您的评论对其它用户来说具有重要的参考价值，所以请认真填写。
- 类似“顶”、“沙发”之类没有营养的文字，对勤劳贡献的楼主来说是令人沮丧的反馈信息。
- 相信您也不想看到一排文字/表情墙，所以请不要反馈意义不大的重复字符，也请尽量不要纯表情的回复。
- 提问之前请再仔细看一遍楼主的说明，或许是您遗漏了。
- 请勿到处挖坑绊人、招贴广告。既占空间让人厌烦，又没人会搭理，于人于己都无利。
关于好例子网
本站旨在为广大IT学习爱好者提供一个非营利性互相学习交流分享平台。本站所有资源都可以被免费获取学习研究。本站资源来自网友分享，对搜索内容的合法性不具有预见性、识别性、控制性，仅供学习研究，请务必在下载后24小时内给予删除，不得用于其他任何用途，否则后果自负。基于互联网的特殊性，平台无法对用户传输的作品、信息、内容的权属或合法性、安全性、合规性、真实性、科学性、完整权、有效性等进行实质审查；无论平台是否已进行审查，用户均应自行承担因其传输的作品、信息、内容而可能或已经产生的侵权或权属纠纷等法律责任。本站所有资源不代表本站的观点或立场，基于网友分享，根据中国法律《信息网络传播权保护条例》第二十二与二十三条之规定，若资源存在侵权或相关问题请联系本站客服人员，点此联系我们。关于更多版权及免责申明参见 版权及免责申明
网友评论
我要评论