stata回归分析及roc分析实例（Stata regression analysis and ROC analysis example）

stata回归分析及roc分析实例（Stata regression analysis and ROC analysis example） stata回归分析及roc分析实例(Stata regression analysis and ROC analysis example) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ (r) / _ _ / _ _ _ _ / / _ _ _ _ / _ _ _ / / / / / / _ _ _ _ _ _ / 12.0 copyright 1985 - 2011 statacorp lp statistics / data analysis statacorp 4905 lakeway drive special edition, college station, texas 77845 usa 800 was pc http: / / www.stata.com 979 - 696 - 4600 stata @ stata.com 979 - 696 - 4601 (fax) notes: 1. (v # option or set maxvar) 5000 maximum variable . use "c: \ program files \ stata12 \ 2 _ data \ 002 - 胰腺 炎.dta", clear . list ik + + | ldh cr |) mods | | nu 1. | 299.3 47.1 34.4 1 | 2. | 2036 395.1 25.9 1 | 3. | 881 89.4 39.1 1 | 4. | 2250 360.2 26.2 1 | 5. | 300 775 22.4 1 | | | nu 6. | 410 177 21.1 1 | 7. | 2047.4 276 27.1 1 | 8. | 633 235.4 24.7 1 | 9. | 225 71 30.2 1 | 10. | 1149 37 21 1 | | | nu 11. | 881 310 34 1 | 12. | 500 318 28 1 | 13. | 1248 180 29 1 | 14. | 345 210 32.7 1 | 15. | 633 235 24.7 1 | | | nu 16. | 460.5 157 26 1 | 17. | 359 159 35.4 1 | 18. | 305 138 17.9 1 | 19. | 1487 306 27 1 | 20. | 1211 205 27.2 1 | | | nu 21. | 407 60 33.4 1 | 22. | 1100 38 for 1 | 23. | 506 92.7 40.9 1 | 24. | 320 107 26.4 1 | 25. | 823 63 1 | 18.8 | | nu 26. | 1402.3 318 28. 2 1 | 27。| 2272 383 21.6 1 | 28。| 489.7 71 36.2 1 | 29。| 245 123 36.2 0 | 30。| 230 21 24.3 0 | | ------------------------------ | 31。| 264 51 30.9 0 | 32。| 293 38 28.8 0 | 33。| 168 52 28.7 0 | 34。| 168 45 33.2 0 | 35。| 224 56 28.2 0 | | ------------------------------ | 36。| 175 78 45 0 | 37。| 290 87 32 0 | 38。| 117 66 38.8 0 | 39。| 147 58 39.7 0 | 40。| 203 51 46.9 0 | | ------------------------------ | 41。| 189 84 41.6 0 | 42。| 658 205 34.4 0 | 43。| 210 46 41.3 0 | 44。| 555 52 31.8 0 | 45。| 264 58 41.8 0 | | ------------------------------ | 46。| 214 79 41 0 | 47。| 181 70 36.5 0 | 48。| 454 80 33.2 0 | 49。| 319 67 38.3 0 | 50。| 1033 88 32.2 0 | | ------------------------------ | 51。| 320 64 31.6 0 | 52。| 171 73 37.2 0 | 53。| 449 59 37.7 0 | 54。| 278 55 35.2 0 | 55。| 2.9 54 37.9 0 | | ------------------------------ | 56。| 354 30 37.9 0 | 57。| 138 78 34.8 0 | 58。| 225 53 36.1 0 | 59。| 323 95 33.1 0 | 60。| 146 87 36.5 0 | | ------------------------------ | 61。| 205 66 33.1 0 | 62。| 174 94 41 0 | 63。| 170 72 35.2 0 | 64。| 270 58 33.9 0 | 65。| 160 63 35.2 0 | | ------------------------------ | 66。| 194 57 32.7 0 | 67。| 158 94 34.5 0 | 68。| 217 50 36.6 0 | 69。| 310 76 27。 7 0 | 70。| 29 40 29 0 | | ------------------------------ | 71。| 186 71 31 0 | 72。| 144 61 34.8 0 | 73。| 145 90 47.6 0 | 74。| 227 61 37.3 0 | 75。| 279 63 45.3 0 | | ------------------------------ | 76。| 148 64 35.6 0 | 77。| 269 76 40.8 0 | 78。| 654 74 36 0 | 79。| 236 74 44 0 | 80。| 203 60 37.7 0 | | ------------------------------ | 81。| 1194 204 38.1 0 | 82。| 982 50 30.4 0 | 83。| 485 83 41.8 0 | 84。| 258 212 31 0 | 85。| 290.3 80 39.7 0 | | ------------------------------ | 86。| 265 73 51.2 0 | 87。| 337 48 35.5 0 | 88。| 218 74 43.8 0 | 89。| 498 101 39.1 0 | 90。| 187 74 28.5 0 | | ------------------------------ | 91。| 186 69 42.3 0 | 92。| 242 90 37.9 0 | 93。| 282 54 41.7 0 | 94。| 144 49 30 0 | 95。| 124 48 42 0 | | ------------------------------ | 96。| 202 70 37.7 0 | 97。| 303 70 33 0 | 98。| 234 61 30.5 0 | 99。| 271 71 40.9 0 | 100。| 157 87 47.7 0 | | ------------------------------ | 101。| 209 74 31 0 | 102。| 168 94 27.6 0 | 103。| 193 45 30.8 0 | 104。| 219 80 35.9 0 | 105。| 131 79 43.5 0 | | ------------------------------ | 106。| 179 61 41 0 | 107。| 235 45 35.6 0 | 108。| 115 71 44.9 0 | 109。| 134 78 39.7 0 | 110。| 369 73 39.2 0 | | ------------------------------ | 111。| 967 81 34。 8 0 | 112。| 184 89 39.7 0 | 113。| 662 96 26.4 0 | ------------------------------ + + 。总和 变| OBS平均标准开发。最小最大 ------------- + -------------------------------------------------------- 113 433.5434 448.6421 2.9 2272 | LDH 113 106.3265 100.756 21 775 | Cr 113 34.45221 6.624105 17.9 51.2 | ABL MODS | 113. 2477876。4336509 0 1 。标签插件 MODS |频率成暨。 ------------ + ----------------------------------- 0 | 85 75.22 75.22 1 | 28 24.78 100 ------------ + ----------------------------------- 总| 113 100 。急性肾功能衰竭 迭代0:对数似然= -63.26774 迭代1:对数似然= -31.249401 迭代2:对数似然= -30.061031 迭代3:对数似然= -30.03929 迭代4:对数似然= -30.039258 迭代5:对数似然= -30.039258 OBS的logistic回归数,113 LRχ(3)= 66.4 问题>χ= 对数似然= -30.039258伪R2 = 0.5252 ----------------------------------------------------------- ------------------- MODS |系数。标准差。Z P,| Z | [ 95% ]设置间隔 ------------- + ---------------------------------------------------------------- LDH |。0024992。001073 2.33 0.020 0046021 0003962。 铬|。0143511。0057272 2.51 0.012 0255761 0031261。 ABL | -。1858638。0647696 -2.87 0.004。3128099。0589177 _. | 2.24286 2.246818 0.318 -2.160823 6.646544? ------------------------------------------------------------------------------ . the logit. logistic regression number of obs = 113 the chi square (2) = 66.46 prob > chi2 = 0.0000 the log likelihood is -30.039258 pseudo r2 = 0.5252 ------------------------------------------------------------------------------ mods | odds ratio std. err. z z z z z z z | (95%) | conf. interval) ----------------- + ----------------------------------------- ldh | 1.002502 0.020 1.000396 1.004613.0010756 2.33. vc | 1.014455.0058099 2.51 0.012 1.003131 1.025906 |.8303867 abl.0537838.731389.9427843 -2.87 0.004 _. | 9.420238 21.16557 0.318.1152303 770.118? ------------------------------------------------------------------------------ . lroc logistic model for mods a number of observations. area under roc curve = 0.9273 . lsens . lfit, g (10) logistic model for mods, goodness of fit test. table i the probability quantiles of fact) a number of observations. number of groups 10 hosmer lemeshow chi2 = 5.93 (8) 0.6549 prob > chi2 = . pre ss (optional) assumed, pr (mods) . roccomp mods, cr ldh abl roc asymptotic normal. obs area std. err. [95% conf. interval) ----------------------------------------------------------- -------------- 0.9273 0.0268 0.87485 0.97977 ss. ldh (0.9034 0. 0285 0.84752 0.95921 CR 113 0.7998 0.0633 0.67580 0.92378 ABL 113 0.1483 0.0444 0.06136 0.23528 ----------------------------------------------------------- -------------- 区域(SS)=区域(LDH)=区域(CR)=区域(ABL) χ(3)= 189.39 = 0 > 21问 。rocgold MODS SS LDH Cr ABL ------------------------------------------------------------------------------- ROC Bonferroni 面积标准差。Chi2 DF公关> Chi2 PR > 21 ------------------------------------------------------------------------------- SS(标准)0.9273 0.0268 乳酸脱氢酶0.9034，0.0285，0.6873，1，0.4071，1 CR 0.7998 0.0633 4.9712 1 0.0258 0.0773 ABL 0.1483 0.0444 135.4836 1 0 0 ------------------------------------------------------------------------------- 。roccomp MODS SS LDH Cr ABL，G 。检测LDH，由(MODS) 方差相等的两样本t检验 ------------------------------------------------------------------------------ 集团| OBS均值标准差。Std. Dev. [ 95% ]设置间隔 --------- + -------------------------------------------------------------------- 0 85 282.4729 22.79759 210.1834 237.1374 327.8085 | 1 28 892.15 120.6442 638.3891 644.6085 1139.691 | --------- + -------------------------------------------------------------------- 结合113 433.5434 42.2047 448.6421 349.9202 517.1666 | --------- + -------------------------------------------------------------------- 不同的| -609.6771 79.33441 -766.8835 -452.4706 ------------------------------------------------------------------------------ 差异=平均(0)均值(1)T = -7.6849 差= 0自由度= 111 哈:差异0～= 0公顷:差异> 0 PR(t,t)= 0，PR(| T | > | T |)= 0 PR(T,t)= 1 。sktest LDH 偏度/峰度检验 -------联合------ 变| OBS PR(偏斜度)PR(峰度)的χ(2)问题> 2 ------------- + ----------------------------------------------------------- ---- 113 0 0 52.62 0 | LDH 。秩和LDH，用(MODS) 两样本比较的秩和检验(Mann Whitney) MODS | OBS秩和预期 ------------- + --------------------------------- 0 | 85 3885 4845 1 | 28 2556 1596 ------------- + --------------------------------- 联合| 113 6441 6441 未经调整的方差为22610 领带-1.13调整 ---------- 调整后的方差为22608.87 病人:乳酸脱氢酶(MODS = 0)=乳酸脱氢酶(MODS = 1) Z = -6.385 问题> | Z | = 0 。