# 2种方式选择1个
install.packages("table1")
require(devtools)
devtools::install_github("benjaminrich/table1")47 table1绘制三线表
三线表/基线资料表作为临床医学论文中最常见的第一个三线表，几乎是100%必备技能！一般我们会通过Word和spss结合进行，在Word里面把三线表画好，然后把数据复制粘贴到Word里。
如果通过R语言的话那么实现方式非常多，前面给大家介绍了compareGroups和tableone两个R包，今天再给大家介绍一下table1。
47.1 安装
47.2 使用
为了方便比较，还是使用compareGroups包自带的regicor数据集。
为了说明这个软件包是如何工作的，我们从REGICOR研究中取了一部分数据。REGICOR是一个 对来自西班牙东北部的参与者进行的横断面研究，包括：人口统计学信息（年龄、性别、身高、体重、腰围等）、血脂特征（总胆固醇和胆固醇、甘油三酯等）、问卷调查信息（体育活动，生活质量，…）等。此外，心血管事件和 死亡信息来自医院和官方登记处。
library(compareGroups)
library(table1)
##
## Attaching package: 'table1'
## The following objects are masked from 'package:base':
##
## units, units<-
data("regicor")
dim(regicor)
## [1] 2294 25
str(regicor)
## 'data.frame': 2294 obs. of 25 variables:
## $ id : num 2.26e+03 1.88e+03 3.00e+09 3.00e+09 3.00e+09 ...
## ..- attr(*, "label")= Named chr "Individual id"
## .. ..- attr(*, "names")= chr "id"
## $ year : Factor w/ 3 levels "1995","2000",..: 3 3 2 2 2 2 2 1 3 1 ...
## ..- attr(*, "label")= Named chr "Recruitment year"
## .. ..- attr(*, "names")= chr "year"
## $ age : int 70 56 37 69 70 40 66 53 43 70 ...
## ..- attr(*, "label")= Named chr "Age"
## .. ..- attr(*, "names")= chr "age"
## $ sex : Factor w/ 2 levels "Male","Female": 2 2 1 2 2 2 1 2 2 1 ...
## ..- attr(*, "label")= chr "Sex"
## $ smoker : Factor w/ 3 levels "Never smoker",..: 1 1 2 1 NA 2 1 1 3 3 ...
## ..- attr(*, "label")= Named chr "Smoking status"
## .. ..- attr(*, "names")= chr "smoker"
## $ sbp : int 138 139 132 168 NA 108 120 132 95 142 ...
## ..- attr(*, "label")= Named chr "Systolic blood pressure"
## .. ..- attr(*, "names")= chr "sbp"
## $ dbp : int 75 89 82 97 NA 70 72 78 65 78 ...
## ..- attr(*, "label")= Named chr "Diastolic blood pressure"
## .. ..- attr(*, "names")= chr "dbp"
## $ histhtn : Factor w/ 2 levels "Yes","No": 2 2 2 2 2 2 1 2 2 2 ...
## ..- attr(*, "label")= Named chr "History of hypertension"
## .. ..- attr(*, "names")= chr "histbp"
## $ txhtn : Factor w/ 2 levels "No","Yes": 1 1 1 1 1 1 2 1 1 1 ...
## ..- attr(*, "label")= chr "Hypertension treatment"
## $ chol : num 294 220 245 168 NA NA 298 254 194 188 ...
## ..- attr(*, "label")= Named chr "Total cholesterol"
## .. ..- attr(*, "names")= chr "chol"
## $ hdl : num 57 50 59.8 53.2 NA ...
## ..- attr(*, "label")= Named chr "HDL cholesterol"
## .. ..- attr(*, "names")= chr "hdl"
## $ triglyc : num 93 160 89 116 NA 94 71 NA 68 137 ...
## ..- attr(*, "label")= Named chr "Triglycerides"
## .. ..- attr(*, "names")= chr "triglyc"
## $ ldl : num 218.4 138 167.4 91.6 NA ...
## ..- attr(*, "label")= Named chr "LDL cholesterol"
## .. ..- attr(*, "names")= chr "ldl"
## $ histchol: Factor w/ 2 levels "Yes","No": 2 2 2 2 NA 2 1 2 2 2 ...
## ..- attr(*, "label")= chr "History of hyperchol."
## $ txchol : Factor w/ 2 levels "No","Yes": 1 1 1 1 NA 1 1 1 1 1 ...
## ..- attr(*, "label")= Named chr "Cholesterol treatment"
## .. ..- attr(*, "names")= chr "txchol"
## $ height : num 160 163 170 147 NA ...
## ..- attr(*, "label")= Named chr "Height (cm)"
## .. ..- attr(*, "names")= chr "height"
## $ weight : num 64 67 70 68 NA 43.5 79.2 45.8 53 62 ...
## ..- attr(*, "label")= Named chr "Weight (Kg)"
## .. ..- attr(*, "names")= chr "weight"
## $ bmi : num 25 25.2 24.2 31.5 NA ...
## ..- attr(*, "label")= Named chr "Body mass index"
## .. ..- attr(*, "names")= chr "bmi"
## $ phyact : num 304 160 553 522 NA ...
## ..- attr(*, "label")= Named chr "Physical activity (Kcal/week)"
## .. ..- attr(*, "names")= chr "phyact"
## $ pcs : num 54.5 58.2 43.4 54.3 NA ...
## ..- attr(*, "label")= Named chr "Physical component"
## .. ..- attr(*, "names")= chr "pcs"
## $ mcs : num 58.9 48 62.6 57.9 NA ...
## ..- attr(*, "label")= chr "Mental component"
## $ cv : Factor w/ 2 levels "No","Yes": 1 1 1 1 NA 1 1 1 1 1 ...
## ..- attr(*, "label")= chr "Cardiovascular event"
## $ tocv : num 1025 2757 1906 1055 NA ...
## ..- attr(*, "label")= chr "Days to cardiovascular event or end of follow-up"
## $ death : Factor w/ 2 levels "No","Yes": 2 1 1 1 NA 1 2 1 1 1 ...
## ..- attr(*, "label")= chr "Overall death"
## $ todeath : num 1299.2 39.3 858.4 1833.1 NA ...
## ..- attr(*, "label")= chr "Days to overall death or end of follow-up"各个变量的信息如下：
| Name | Label | Codes |
|---|---|---|
| id | Individual id | |
| year | Recruitment year | 1995; 2000; 2005 |
| age | Age | |
| sex | Sex | Male; Female |
| smoker | Smoking status | Never smoker; Current or former < 1y; Former $\geq$ 1y |
| sbp | Systolic blood pressure | |
| dbp | Diastolic blood pressure | |
| histhtn | History of hypertension | Yes; No |
| txhtn | Hypertension treatment | No; Yes |
| chol | Total cholesterol | |
| hdl | HDL cholesterol | |
| triglyc | Triglycerides | |
| ldl | LDL cholesterol | |
| histchol | History of hyperchol. | Yes; No |
| txchol | Cholesterol treatment | No; Yes |
| height | Height (cm) | |
| weight | Weight (Kg) | |
| bmi | Body mass index | |
| phyact | Physical activity (Kcal/week) | |
| pcs | Physical component | |
| mcs | Mental component | |
| cv | Cardiovascular event | No; Yes |
| tocv | Days to cardiovascular event or end of follow-up | |
| death | Overall death | No; Yes |
| todeath | Days to overall death or end of follow-up |
大家在使用的时候还是要注意，要先把分类变量因子化！使用起来也是1行代码就可以出图了：
table1(~ . | year, data = regicor[,-1])| 1995 (N=431) |
2000 (N=786) |
2005 (N=1077) |
Overall (N=2294) |
|
|---|---|---|---|---|
| Age | ||||
| Mean (SD) | 54.1 (11.7) | 54.3 (11.2) | 55.3 (10.6) | 54.7 (11.0) |
| Median [Min, Max] | 54.0 [35.0, 74.0] | 54.0 [35.0, 74.0] | 56.0 [35.0, 74.0] | 55.0 [35.0, 74.0] |
| Sex | ||||
| Male | 206 (47.8%) | 390 (49.6%) | 505 (46.9%) | 1101 (48.0%) |
| Female | 225 (52.2%) | 396 (50.4%) | 572 (53.1%) | 1193 (52.0%) |
| Smoking status | ||||
| Never smoker | 234 (54.3%) | 414 (52.7%) | 553 (51.3%) | 1201 (52.4%) |
| Current or former < 1y | 109 (25.3%) | 267 (34.0%) | 217 (20.1%) | 593 (25.9%) |
| Former >= 1y | 72 (16.7%) | 77 (9.8%) | 290 (26.9%) | 439 (19.1%) |
| Missing | 16 (3.7%) | 28 (3.6%) | 17 (1.6%) | 61 (2.7%) |
| Systolic blood pressure | ||||
| Mean (SD) | 133 (19.2) | 133 (21.3) | 129 (19.8) | 131 (20.3) |
| Median [Min, Max] | 132 [92.0, 196] | 132 [80.0, 212] | 125 [88.0, 229] | 129 [80.0, 229] |
| Missing | 3 (0.7%) | 11 (1.4%) | 0 (0%) | 14 (0.6%) |
| Diastolic blood pressure | ||||
| Mean (SD) | 77.0 (10.5) | 80.8 (10.3) | 79.9 (10.6) | 79.7 (10.5) |
| Median [Min, Max] | 78.0 [52.0, 108] | 80.0 [50.0, 123] | 79.0 [40.0, 121] | 80.0 [40.0, 123] |
| Missing | 3 (0.7%) | 11 (1.4%) | 0 (0%) | 14 (0.6%) |
| History of hypertension | ||||
| Yes | 111 (25.8%) | 233 (29.6%) | 379 (35.2%) | 723 (31.5%) |
| No | 320 (74.2%) | 553 (70.4%) | 690 (64.1%) | 1563 (68.1%) |
| Missing | 0 (0%) | 0 (0%) | 8 (0.7%) | 8 (0.3%) |
| Hypertension treatment | ||||
| No | 360 (83.5%) | 659 (83.8%) | 804 (74.7%) | 1823 (79.5%) |
| Yes | 71 (16.5%) | 127 (16.2%) | 230 (21.4%) | 428 (18.7%) |
| Missing | 0 (0%) | 0 (0%) | 43 (4.0%) | 43 (1.9%) |
| Total cholesterol | ||||
| Mean (SD) | 225 (43.1) | 224 (44.4) | 213 (45.9) | 219 (45.2) |
| Median [Min, Max] | 225 [95.0, 361] | 222 [113, 404] | 209 [95.0, 488] | 215 [95.0, 488] |
| Missing | 28 (6.5%) | 71 (9.0%) | 2 (0.2%) | 101 (4.4%) |
| HDL cholesterol | ||||
| Mean (SD) | 51.9 (14.5) | 52.3 (15.6) | 53.2 (14.2) | 52.7 (14.7) |
| Median [Min, Max] | 50.6 [24.5, 95.0] | 49.7 [19.6, 103] | 52.0 [20.0, 112] | 51.0 [19.6, 112] |
| Missing | 30 (7.0%) | 38 (4.8%) | 1 (0.1%) | 69 (3.0%) |
| Triglycerides | ||||
| Mean (SD) | 114 (74.4) | 114 (70.7) | 117 (76.0) | 116 (73.9) |
| Median [Min, Max] | 94.0 [41.0, 882] | 98.0 [25.0, 960] | 98.0 [33.0, 744] | 97.0 [25.0, 960] |
| Missing | 28 (6.5%) | 34 (4.3%) | 1 (0.1%) | 63 (2.7%) |
| LDL cholesterol | ||||
| Mean (SD) | 152 (38.4) | 149 (38.6) | 136 (39.7) | 143 (39.7) |
| Median [Min, Max] | 151 [36.3, 273] | 147 [51.7, 300] | 133 [36.8, 330] | 141 [36.3, 330] |
| Missing | 43 (10.0%) | 98 (12.5%) | 27 (2.5%) | 168 (7.3%) |
| History of hyperchol. | ||||
| Yes | 97 (22.5%) | 256 (32.6%) | 356 (33.1%) | 709 (30.9%) |
| No | 334 (77.5%) | 515 (65.5%) | 715 (66.4%) | 1564 (68.2%) |
| Missing | 0 (0%) | 15 (1.9%) | 6 (0.6%) | 21 (0.9%) |
| Cholesterol treatment | ||||
| No | 403 (93.5%) | 705 (89.7%) | 903 (83.8%) | 2011 (87.7%) |
| Yes | 28 (6.5%) | 68 (8.7%) | 132 (12.3%) | 228 (9.9%) |
| Missing | 0 (0%) | 13 (1.7%) | 42 (3.9%) | 55 (2.4%) |
| Height (cm) | ||||
| Mean (SD) | 163 (9.21) | 162 (9.39) | 163 (9.05) | 163 (9.22) |
| Median [Min, Max] | 163 [143, 192] | 162 [139, 190] | 163 [137, 199] | 163 [137, 199] |
| Missing | 8 (1.9%) | 15 (1.9%) | 12 (1.1%) | 35 (1.5%) |
| Weight (Kg) | ||||
| Mean (SD) | 72.3 (12.6) | 73.8 (14.0) | 73.6 (13.9) | 73.4 (13.7) |
| Median [Min, Max] | 72.0 [41.2, 127] | 73.0 [43.5, 118] | 73.0 [42.0, 125] | 73.0 [41.2, 127] |
| Missing | 8 (1.9%) | 15 (1.9%) | 12 (1.1%) | 35 (1.5%) |
| Body mass index | ||||
| Mean (SD) | 27.0 (4.15) | 28.1 (4.62) | 27.6 (4.63) | 27.6 (4.56) |
| Median [Min, Max] | 26.5 [17.2, 43.7] | 27.6 [17.4, 48.2] | 27.2 [17.1, 48.2] | 27.2 [17.1, 48.2] |
| Missing | 8 (1.9%) | 15 (1.9%) | 12 (1.1%) | 35 (1.5%) |
| Physical activity (Kcal/week) | ||||
| Mean (SD) | 491 (419) | 422 (377) | 351 (378) | 399 (388) |
| Median [Min, Max] | 390 [1.00, 3430] | 347 [0, 5080] | 262 [0, 4230] | 304 [0, 5080] |
| Missing | 64 (14.8%) | 22 (2.8%) | 2 (0.2%) | 88 (3.8%) |
| Physical component | ||||
| Mean (SD) | 49.3 (8.08) | 49.0 (9.63) | 50.1 (8.91) | 49.6 (9.01) |
| Median [Min, Max] | 51.7 [18.6, 66.2] | 51.9 [13.9, 67.0] | 53.1 [17.9, 67.1] | 52.3 [13.9, 67.1] |
| Missing | 34 (7.9%) | 123 (15.6%) | 83 (7.7%) | 240 (10.5%) |
| Mental component | ||||
| Mean (SD) | 49.2 (11.3) | 48.9 (11.0) | 46.9 (10.8) | 48.0 (11.0) |
| Median [Min, Max] | 52.3 [3.42, 69.9] | 52.8 [10.2, 69.4] | 49.7 [7.62, 67.6] | 51.3 [3.42, 69.9] |
| Missing | 34 (7.9%) | 123 (15.6%) | 83 (7.7%) | 240 (10.5%) |
| Cardiovascular event | ||||
| No | 388 (90.0%) | 706 (89.8%) | 977 (90.7%) | 2071 (90.3%) |
| Yes | 10 (2.3%) | 35 (4.5%) | 47 (4.4%) | 92 (4.0%) |
| Missing | 33 (7.7%) | 45 (5.7%) | 53 (4.9%) | 131 (5.7%) |
| Days to cardiovascular event or end of follow-up | ||||
| Mean (SD) | 1780 (1100) | 1690 (1080) | 1790 (1070) | 1750 (1080) |
| Median [Min, Max] | 1730 [0.260, 3650] | 1620 [0.115, 3640] | 1770 [2.65, 3650] | 1720 [0.115, 3650] |
| Missing | 33 (7.7%) | 45 (5.7%) | 53 (4.9%) | 131 (5.7%) |
| Overall death | ||||
| No | 369 (85.6%) | 657 (83.6%) | 949 (88.1%) | 1975 (86.1%) |
| Yes | 18 (4.2%) | 81 (10.3%) | 74 (6.9%) | 173 (7.5%) |
| Missing | 44 (10.2%) | 48 (6.1%) | 54 (5.0%) | 146 (6.4%) |
| Days to overall death or end of follow-up | ||||
| Mean (SD) | 1710 (1040) | 1670 (1050) | 1760 (1050) | 1720 (1050) |
| Median [Min, Max] | 1560 [0.220, 3640] | 1610 [0.897, 3650] | 1730 [0.889, 3650] | 1670 [0.220, 3650] |
| Missing | 44 (10.2%) | 48 (6.1%) | 54 (5.0%) | 146 (6.4%) |
是不是很简单，把需要分组的变量放在|后面即可。
结果是html形式的，可以直接复制粘贴到Word里面，连续性变量默认给出了两种表示方式！
47.3 美化
先给大家展示下复杂表头的制作格式，其实非常简单，就是用*即可。
table1(~ . | year*sex, data = regicor[,-1],overall = F) # 不展示overall
1995
|
2000
|
2005
|
||||
|---|---|---|---|---|---|---|
| Male (N=206) |
Female (N=225) |
Male (N=390) |
Female (N=396) |
Male (N=505) |
Female (N=572) |
|
| Age | ||||||
| Mean (SD) | 54.1 (11.8) | 54.1 (11.7) | 54.3 (11.2) | 54.4 (11.2) | 55.4 (10.7) | 55.2 (10.6) |
| Median [Min, Max] | 53.0 [35.0, 74.0] | 54.0 [35.0, 74.0] | 54.0 [35.0, 74.0] | 54.0 [35.0, 74.0] | 55.0 [35.0, 74.0] | 56.0 [35.0, 74.0] |
| Smoking status | ||||||
| Never smoker | 52 (25.2%) | 182 (80.9%) | 112 (28.7%) | 302 (76.3%) | 137 (27.1%) | 416 (72.7%) |
| Current or former < 1y | 77 (37.4%) | 32 (14.2%) | 199 (51.0%) | 68 (17.2%) | 134 (26.5%) | 83 (14.5%) |
| Former >= 1y | 67 (32.5%) | 5 (2.2%) | 66 (16.9%) | 11 (2.8%) | 227 (45.0%) | 63 (11.0%) |
| Missing | 10 (4.9%) | 6 (2.7%) | 13 (3.3%) | 15 (3.8%) | 7 (1.4%) | 10 (1.7%) |
| Systolic blood pressure | ||||||
| Mean (SD) | 134 (18.4) | 132 (19.8) | 137 (19.3) | 129 (22.6) | 132 (18.7) | 127 (20.5) |
| Median [Min, Max] | 132 [92.0, 184] | 132 [92.0, 196] | 134 [90.0, 212] | 124 [80.0, 206] | 129 [93.0, 200] | 122 [88.0, 229] |
| Missing | 0 (0%) | 3 (1.3%) | 3 (0.8%) | 8 (2.0%) | 0 (0%) | 0 (0%) |
| Diastolic blood pressure | ||||||
| Mean (SD) | 79.0 (9.27) | 75.2 (11.3) | 83.0 (9.54) | 78.6 (10.6) | 81.7 (10.8) | 78.3 (10.0) |
| Median [Min, Max] | 78.0 [54.0, 108] | 74.0 [52.0, 102] | 82.0 [60.0, 123] | 80.0 [50.0, 112] | 81.0 [40.0, 121] | 78.0 [47.0, 111] |
| Missing | 0 (0%) | 3 (1.3%) | 3 (0.8%) | 8 (2.0%) | 0 (0%) | 0 (0%) |
| History of hypertension | ||||||
| Yes | 50 (24.3%) | 61 (27.1%) | 110 (28.2%) | 123 (31.1%) | 181 (35.8%) | 198 (34.6%) |
| No | 156 (75.7%) | 164 (72.9%) | 280 (71.8%) | 273 (68.9%) | 319 (63.2%) | 371 (64.9%) |
| Missing | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 5 (1.0%) | 3 (0.5%) |
| Hypertension treatment | ||||||
| No | 175 (85.0%) | 185 (82.2%) | 342 (87.7%) | 317 (80.1%) | 372 (73.7%) | 432 (75.5%) |
| Yes | 31 (15.0%) | 40 (17.8%) | 48 (12.3%) | 79 (19.9%) | 110 (21.8%) | 120 (21.0%) |
| Missing | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 23 (4.6%) | 20 (3.5%) |
| Total cholesterol | ||||||
| Mean (SD) | 224 (43.9) | 226 (42.4) | 224 (43.9) | 224 (44.9) | 210 (40.3) | 216 (50.3) |
| Median [Min, Max] | 224 [101, 346] | 227 [95.0, 361] | 226 [122, 404] | 220 [113, 386] | 207 [95.0, 369] | 211 [109, 488] |
| Missing | 10 (4.9%) | 18 (8.0%) | 37 (9.5%) | 34 (8.6%) | 0 (0%) | 2 (0.3%) |
| HDL cholesterol | ||||||
| Mean (SD) | 46.5 (13.1) | 56.9 (13.9) | 47.3 (12.6) | 57.4 (16.7) | 48.1 (12.4) | 57.8 (14.2) |
| Median [Min, Max] | 45.3 [24.5, 95.0] | 56.6 [28.6, 92.6] | 45.7 [19.6, 94.2] | 56.8 [20.1, 103] | 47.0 [20.0, 97.0] | 57.0 [21.0, 112] |
| Missing | 11 (5.3%) | 19 (8.4%) | 15 (3.8%) | 23 (5.8%) | 0 (0%) | 1 (0.2%) |
| Triglycerides | ||||||
| Mean (SD) | 131 (91.5) | 97.8 (47.9) | 128 (81.1) | 99.7 (55.1) | 132 (90.3) | 104 (57.6) |
| Median [Min, Max] | 110 [43.0, 882] | 86.0 [41.0, 364] | 113 [25.0, 960] | 87.0 [31.0, 604] | 108 [33.0, 744] | 90.0 [33.0, 521] |
| Missing | 9 (4.4%) | 19 (8.4%) | 17 (4.4%) | 17 (4.3%) | 0 (0%) | 1 (0.2%) |
| LDL cholesterol | ||||||
| Mean (SD) | 153 (39.6) | 150 (37.3) | 152 (39.1) | 146 (38.0) | 137 (36.0) | 136 (42.6) |
| Median [Min, Max] | 153 [36.3, 266] | 148 [38.6, 273] | 150 [51.7, 300] | 142 [57.7, 290] | 132 [38.2, 289] | 133 [36.8, 330] |
| Missing | 18 (8.7%) | 25 (11.1%) | 52 (13.3%) | 46 (11.6%) | 21 (4.2%) | 6 (1.0%) |
| History of hyperchol. | ||||||
| Yes | 48 (23.3%) | 49 (21.8%) | 138 (35.4%) | 118 (29.8%) | 167 (33.1%) | 189 (33.0%) |
| No | 158 (76.7%) | 176 (78.2%) | 247 (63.3%) | 268 (67.7%) | 336 (66.5%) | 379 (66.3%) |
| Missing | 0 (0%) | 0 (0%) | 5 (1.3%) | 10 (2.5%) | 2 (0.4%) | 4 (0.7%) |
| Cholesterol treatment | ||||||
| No | 189 (91.7%) | 214 (95.1%) | 348 (89.2%) | 357 (90.2%) | 425 (84.2%) | 478 (83.6%) |
| Yes | 17 (8.3%) | 11 (4.9%) | 38 (9.7%) | 30 (7.6%) | 59 (11.7%) | 73 (12.8%) |
| Missing | 0 (0%) | 0 (0%) | 4 (1.0%) | 9 (2.3%) | 21 (4.2%) | 21 (3.7%) |
| Height (cm) | ||||||
| Mean (SD) | 170 (7.34) | 158 (6.31) | 168 (7.17) | 156 (6.50) | 170 (7.43) | 158 (6.24) |
| Median [Min, Max] | 170 [148, 192] | 158 [143, 178] | 168 [144, 190] | 155 [139, 176] | 170 [148, 199] | 157 [137, 180] |
| Missing | 2 (1.0%) | 6 (2.7%) | 5 (1.3%) | 10 (2.5%) | 4 (0.8%) | 8 (1.4%) |
| Weight (Kg) | ||||||
| Mean (SD) | 77.6 (11.7) | 67.3 (11.3) | 80.1 (12.3) | 67.6 (12.6) | 80.2 (11.6) | 67.7 (13.0) |
| Median [Min, Max] | 77.3 [41.2, 127] | 66.0 [43.2, 98.6] | 79.2 [50.0, 117] | 66.0 [43.5, 118] | 79.0 [43.0, 122] | 66.0 [42.0, 125] |
| Missing | 2 (1.0%) | 6 (2.7%) | 5 (1.3%) | 10 (2.5%) | 4 (0.8%) | 8 (1.4%) |
| Body mass index | ||||||
| Mean (SD) | 26.9 (3.64) | 27.2 (4.57) | 28.2 (3.89) | 28.0 (5.25) | 27.9 (3.58) | 27.3 (5.39) |
| Median [Min, Max] | 26.5 [18.5, 39.6] | 26.4 [17.2, 43.7] | 28.0 [18.4, 40.5] | 27.0 [17.4, 48.2] | 27.7 [18.1, 42.6] | 26.7 [17.1, 48.2] |
| Missing | 2 (1.0%) | 6 (2.7%) | 5 (1.3%) | 10 (2.5%) | 4 (0.8%) | 8 (1.4%) |
| Physical activity (Kcal/week) | ||||||
| Mean (SD) | 422 (418) | 553 (412) | 356 (362) | 486 (382) | 439 (467) | 273 (253) |
| Median [Min, Max] | 291 [1.00, 3240] | 479 [18.0, 3430] | 264 [0, 4220] | 417 [0, 5080] | 317 [0, 4230] | 222 [0, 1960] |
| Missing | 31 (15.0%) | 33 (14.7%) | 10 (2.6%) | 12 (3.0%) | 0 (0%) | 2 (0.3%) |
| Physical component | ||||||
| Mean (SD) | 50.1 (6.71) | 48.6 (9.16) | 50.9 (8.58) | 47.1 (10.2) | 51.5 (8.07) | 48.9 (9.45) |
| Median [Min, Max] | 51.9 [23.3, 65.9] | 51.2 [18.6, 66.2] | 54.3 [13.9, 67.0] | 49.7 [18.3, 65.1] | 54.5 [18.2, 67.1] | 51.6 [17.9, 64.6] |
| Missing | 12 (5.8%) | 22 (9.8%) | 57 (14.6%) | 66 (16.7%) | 30 (5.9%) | 53 (9.3%) |
| Mental component | ||||||
| Mean (SD) | 52.1 (9.67) | 46.5 (12.2) | 50.9 (10.2) | 46.9 (11.3) | 49.2 (9.67) | 44.7 (11.2) |
| Median [Min, Max] | 54.4 [3.42, 66.4] | 49.4 [9.93, 69.9] | 54.5 [14.5, 65.4] | 49.6 [10.2, 69.4] | 51.9 [7.62, 67.6] | 47.3 [12.5, 65.1] |
| Missing | 12 (5.8%) | 22 (9.8%) | 57 (14.6%) | 66 (16.7%) | 30 (5.9%) | 53 (9.3%) |
| Cardiovascular event | ||||||
| No | 190 (92.2%) | 198 (88.0%) | 345 (88.5%) | 361 (91.2%) | 461 (91.3%) | 516 (90.2%) |
| Yes | 6 (2.9%) | 4 (1.8%) | 21 (5.4%) | 14 (3.5%) | 19 (3.8%) | 28 (4.9%) |
| Missing | 10 (4.9%) | 23 (10.2%) | 24 (6.2%) | 21 (5.3%) | 25 (5.0%) | 28 (4.9%) |
| Days to cardiovascular event or end of follow-up | ||||||
| Mean (SD) | 1720 (1130) | 1850 (1070) | 1650 (1080) | 1720 (1080) | 1830 (1060) | 1760 (1080) |
| Median [Min, Max] | 1620 [0.260, 3640] | 1830 [26.4, 3650] | 1610 [0.115, 3640] | 1620 [0.279, 3630] | 1820 [2.65, 3640] | 1750 [7.22, 3650] |
| Missing | 10 (4.9%) | 23 (10.2%) | 24 (6.2%) | 21 (5.3%) | 25 (5.0%) | 28 (4.9%) |
| Overall death | ||||||
| No | 174 (84.5%) | 195 (86.7%) | 321 (82.3%) | 336 (84.8%) | 448 (88.7%) | 501 (87.6%) |
| Yes | 12 (5.8%) | 6 (2.7%) | 46 (11.8%) | 35 (8.8%) | 29 (5.7%) | 45 (7.9%) |
| Missing | 20 (9.7%) | 24 (10.7%) | 23 (5.9%) | 25 (6.3%) | 28 (5.5%) | 26 (4.5%) |
| Days to overall death or end of follow-up | ||||||
| Mean (SD) | 1690 (1030) | 1740 (1050) | 1660 (1030) | 1680 (1070) | 1680 (1030) | 1820 (1070) |
| Median [Min, Max] | 1560 [0.220, 3640] | 1590 [47.7, 3640] | 1610 [0.897, 3620] | 1620 [1.78, 3650] | 1690 [0.889, 3630] | 1810 [1.05, 3650] |
| Missing | 20 (9.7%) | 24 (10.7%) | 23 (5.9%) | 25 (6.3%) | 28 (5.5%) | 26 (4.5%) |
还可以制作更加复杂的表头，比如把2000年和2005年放在同一个表头下，1995年单独放。
这种情况，就需要先进行一点自定义设置了。
labels <- list(
# 先设置需要展示的变量,并设置下展示的名字
variables = list(sex="Sex",age="Age(year)",smoker="Smoking status",
sbp="Systolic blood pressure",dbp="Diastolic blood pressure",
histhtn="History of hypertension",
txhtn="Hypertension treatment",chol="Total cholesterol",
hdl="HDL cholesterol",triglyc="Triglycerides",
ldl="LDL cholesterol"
),
# 设置分组
groups = list("","","2000年以后")
)
# 因子化分组变量，但regicor数据集已经因子化了，这里不用也可以
levels(regicor$year) <- c("1995","2000","2005")然后再设置一下分组变量：
strata <- c(list(Total=regicor[,-1]),split(regicor,regicor$year))然后就能愉快的显示了，是不是很强？
table1(strata, labels, groupspan = c(1,1,2),
render.continuous=c(.="Mean ± SD") # 设置连续性变量的显示方式
)
2000年以后
|
||||
|---|---|---|---|---|
| Total (N=2294) |
1995 (N=431) |
2000 (N=786) |
2005 (N=1077) |
|
| Sex | ||||
| Male | 1101 (48.0%) | 206 (47.8%) | 390 (49.6%) | 505 (46.9%) |
| Female | 1193 (52.0%) | 225 (52.2%) | 396 (50.4%) | 572 (53.1%) |
| Age(year) | ||||
| Mean ± SD | 54.7 ± 11.0 | 54.1 ± 11.7 | 54.3 ± 11.2 | 55.3 ± 10.6 |
| Smoking status | ||||
| Never smoker | 1201 (52.4%) | 234 (54.3%) | 414 (52.7%) | 553 (51.3%) |
| Current or former < 1y | 593 (25.9%) | 109 (25.3%) | 267 (34.0%) | 217 (20.1%) |
| Former >= 1y | 439 (19.1%) | 72 (16.7%) | 77 (9.8%) | 290 (26.9%) |
| Missing | 61 (2.7%) | 16 (3.7%) | 28 (3.6%) | 17 (1.6%) |
| Systolic blood pressure | ||||
| Mean ± SD | 131 ± 20.3 | 133 ± 19.2 | 133 ± 21.3 | 129 ± 19.8 |
| Missing | 14 (0.6%) | 3 (0.7%) | 11 (1.4%) | 0 (0%) |
| Diastolic blood pressure | ||||
| Mean ± SD | 79.7 ± 10.5 | 77.0 ± 10.5 | 80.8 ± 10.3 | 79.9 ± 10.6 |
| Missing | 14 (0.6%) | 3 (0.7%) | 11 (1.4%) | 0 (0%) |
| History of hypertension | ||||
| Yes | 723 (31.5%) | 111 (25.8%) | 233 (29.6%) | 379 (35.2%) |
| No | 1563 (68.1%) | 320 (74.2%) | 553 (70.4%) | 690 (64.1%) |
| Missing | 8 (0.3%) | 0 (0%) | 0 (0%) | 8 (0.7%) |
| Hypertension treatment | ||||
| No | 1823 (79.5%) | 360 (83.5%) | 659 (83.8%) | 804 (74.7%) |
| Yes | 428 (18.7%) | 71 (16.5%) | 127 (16.2%) | 230 (21.4%) |
| Missing | 43 (1.9%) | 0 (0%) | 0 (0%) | 43 (4.0%) |
| Total cholesterol | ||||
| Mean ± SD | 219 ± 45.2 | 225 ± 43.1 | 224 ± 44.4 | 213 ± 45.9 |
| Missing | 101 (4.4%) | 28 (6.5%) | 71 (9.0%) | 2 (0.2%) |
| HDL cholesterol | ||||
| Mean ± SD | 52.7 ± 14.7 | 51.9 ± 14.5 | 52.3 ± 15.6 | 53.2 ± 14.2 |
| Missing | 69 (3.0%) | 30 (7.0%) | 38 (4.8%) | 1 (0.1%) |
| Triglycerides | ||||
| Mean ± SD | 116 ± 73.9 | 114 ± 74.4 | 114 ± 70.7 | 117 ± 76.0 |
| Missing | 63 (2.7%) | 28 (6.5%) | 34 (4.3%) | 1 (0.1%) |
| LDL cholesterol | ||||
| Mean ± SD | 143 ± 39.7 | 152 ± 38.4 | 149 ± 38.6 | 136 ± 39.7 |
| Missing | 168 (7.3%) | 43 (10.0%) | 98 (12.5%) | 27 (2.5%) |
是不是很好看？基本符合要求了！
但是如果有的数值型变量不符合正态分布怎么办？
可以通过自定义函数解决。
# 自定义你的变量要如何显示
rndr <- function(x, name, ...) {
# 分类变量就用默认方法
if (!is.numeric(x)) return(render.categorical.default(x))
# 连续型变量自定义一下
what <- switch(name,
age = "Mean (SD)",
sbp = "Mean (SD)",
dbp = "Mean ± SD",
chol = "Median [IQR]",
hdl = "Median [IQR]",
triglyc = "Median [Min, Max]",
ldl = "Median [IQR]"
)
parse.abbrev.render.code(c("", what))(x)
}这样就能愉快的显示了，不过上面那段代码还是太复杂了~
table1(strata, labels, groupspan = c(1,1,2),
render=rndr
)
2000年以后
|
||||
|---|---|---|---|---|
| Total (N=2294) |
1995 (N=431) |
2000 (N=786) |
2005 (N=1077) |
|
| Sex | ||||
| Male | 1101 (48.0%) | 206 (47.8%) | 390 (49.6%) | 505 (46.9%) |
| Female | 1193 (52.0%) | 225 (52.2%) | 396 (50.4%) | 572 (53.1%) |
| Age(year) | ||||
| Mean (SD) | 54.7 (11.0) | 54.1 (11.7) | 54.3 (11.2) | 55.3 (10.6) |
| Smoking status | ||||
| Never smoker | 1201 (52.4%) | 234 (54.3%) | 414 (52.7%) | 553 (51.3%) |
| Current or former < 1y | 593 (25.9%) | 109 (25.3%) | 267 (34.0%) | 217 (20.1%) |
| Former >= 1y | 439 (19.1%) | 72 (16.7%) | 77 (9.8%) | 290 (26.9%) |
| Systolic blood pressure | ||||
| Mean (SD) | 131 (20.3) | 133 (19.2) | 133 (21.3) | 129 (19.8) |
| Diastolic blood pressure | ||||
| Mean ± SD | 79.7 ± 10.5 | 77.0 ± 10.5 | 80.8 ± 10.3 | 79.9 ± 10.6 |
| History of hypertension | ||||
| Yes | 723 (31.5%) | 111 (25.8%) | 233 (29.6%) | 379 (35.2%) |
| No | 1563 (68.1%) | 320 (74.2%) | 553 (70.4%) | 690 (64.1%) |
| Hypertension treatment | ||||
| No | 1823 (79.5%) | 360 (83.5%) | 659 (83.8%) | 804 (74.7%) |
| Yes | 428 (18.7%) | 71 (16.5%) | 127 (16.2%) | 230 (21.4%) |
| Total cholesterol | ||||
| Median [IQR] | 215 [56.0] | 225 [58.0] | 222 [57.0] | 209 [54.0] |
| HDL cholesterol | ||||
| Median [IQR] | 51.0 [19.4] | 50.6 [20.0] | 49.7 [21.3] | 52.0 [18.0] |
| Triglycerides | ||||
| Median [Min, Max] | 97.0 [25.0, 960] | 94.0 [41.0, 882] | 98.0 [25.0, 960] | 98.0 [33.0, 744] |
| LDL cholesterol | ||||
| Median [IQR] | 141 [52.4] | 151 [48.8] | 147 [53.4] | 133 [48.8] |
还有很多自带的样式可以使用：
table1(strata, labels, groupspan = c(1,1,2),
render=rndr, topclass="Rtable1-zebra"
)
2000年以后
|
||||
|---|---|---|---|---|
| Total (N=2294) |
1995 (N=431) |
2000 (N=786) |
2005 (N=1077) |
|
| Sex | ||||
| Male | 1101 (48.0%) | 206 (47.8%) | 390 (49.6%) | 505 (46.9%) |
| Female | 1193 (52.0%) | 225 (52.2%) | 396 (50.4%) | 572 (53.1%) |
| Age(year) | ||||
| Mean (SD) | 54.7 (11.0) | 54.1 (11.7) | 54.3 (11.2) | 55.3 (10.6) |
| Smoking status | ||||
| Never smoker | 1201 (52.4%) | 234 (54.3%) | 414 (52.7%) | 553 (51.3%) |
| Current or former < 1y | 593 (25.9%) | 109 (25.3%) | 267 (34.0%) | 217 (20.1%) |
| Former >= 1y | 439 (19.1%) | 72 (16.7%) | 77 (9.8%) | 290 (26.9%) |
| Systolic blood pressure | ||||
| Mean (SD) | 131 (20.3) | 133 (19.2) | 133 (21.3) | 129 (19.8) |
| Diastolic blood pressure | ||||
| Mean ± SD | 79.7 ± 10.5 | 77.0 ± 10.5 | 80.8 ± 10.3 | 79.9 ± 10.6 |
| History of hypertension | ||||
| Yes | 723 (31.5%) | 111 (25.8%) | 233 (29.6%) | 379 (35.2%) |
| No | 1563 (68.1%) | 320 (74.2%) | 553 (70.4%) | 690 (64.1%) |
| Hypertension treatment | ||||
| No | 1823 (79.5%) | 360 (83.5%) | 659 (83.8%) | 804 (74.7%) |
| Yes | 428 (18.7%) | 71 (16.5%) | 127 (16.2%) | 230 (21.4%) |
| Total cholesterol | ||||
| Median [IQR] | 215 [56.0] | 225 [58.0] | 222 [57.0] | 209 [54.0] |
| HDL cholesterol | ||||
| Median [IQR] | 51.0 [19.4] | 50.6 [20.0] | 49.7 [21.3] | 52.0 [18.0] |
| Triglycerides | ||||
| Median [Min, Max] | 97.0 [25.0, 960] | 94.0 [41.0, 882] | 98.0 [25.0, 960] | 98.0 [33.0, 744] |
| LDL cholesterol | ||||
| Median [IQR] | 141 [52.4] | 151 [48.8] | 147 [53.4] | 133 [48.8] |
47.4 添加P值
添加P值可以通过自定义函数解决，这样做的好处是可以自定义使用的方法，不好的一点是太复杂了，小白不友好，不符合简单快捷的原则。
# 构建函数，对不同类型自定义统计方法
pvalue <- function(x, ...) {
# Construct vectors of data y, and groups (strata) g
y <- unlist(x)
g <- factor(rep(1:length(x), times=sapply(x, length)))
if (is.numeric(y)) {
# 连续型变量用方差分析
p <- oneway.test(y ~ g)$p.value
} else {
# 分类变量采用卡方检验
p <- chisq.test(table(y, g))$p.value
}
# 设置p值的表示形式
# The initial empty string places the output on the line below the variable label.
c("", sub("<", "<", format.pval(p, digits=3, eps=0.001)))
}但是添加了P值就不能使用自定义的表头了，果然是鱼与熊掌不可兼得啊
table1(~ . | year, data = regicor[,-1], overall = F,
extra.col=list(`P-value`=pvalue),
render.continuous=c(.="Mean ± SD")
)| 1995 (N=431) |
2000 (N=786) |
2005 (N=1077) |
P-value | |
|---|---|---|---|---|
| Age | ||||
| Mean ± SD | 54.1 ± 11.7 | 54.3 ± 11.2 | 55.3 ± 10.6 | 0.0786 |
| Sex | ||||
| Male | 206 (47.8%) | 390 (49.6%) | 505 (46.9%) | 0.506 |
| Female | 225 (52.2%) | 396 (50.4%) | 572 (53.1%) | |
| Smoking status | ||||
| Never smoker | 234 (54.3%) | 414 (52.7%) | 553 (51.3%) | <0.001 |
| Current or former < 1y | 109 (25.3%) | 267 (34.0%) | 217 (20.1%) | |
| Former >= 1y | 72 (16.7%) | 77 (9.8%) | 290 (26.9%) | |
| Missing | 16 (3.7%) | 28 (3.6%) | 17 (1.6%) | |
| Systolic blood pressure | ||||
| Mean ± SD | 133 ± 19.2 | 133 ± 21.3 | 129 ± 19.8 | <0.001 |
| Missing | 3 (0.7%) | 11 (1.4%) | 0 (0%) | |
| Diastolic blood pressure | ||||
| Mean ± SD | 77.0 ± 10.5 | 80.8 ± 10.3 | 79.9 ± 10.6 | <0.001 |
| Missing | 3 (0.7%) | 11 (1.4%) | 0 (0%) | |
| History of hypertension | ||||
| Yes | 111 (25.8%) | 233 (29.6%) | 379 (35.2%) | <0.001 |
| No | 320 (74.2%) | 553 (70.4%) | 690 (64.1%) | |
| Missing | 0 (0%) | 0 (0%) | 8 (0.7%) | |
| Hypertension treatment | ||||
| No | 360 (83.5%) | 659 (83.8%) | 804 (74.7%) | 0.00152 |
| Yes | 71 (16.5%) | 127 (16.2%) | 230 (21.4%) | |
| Missing | 0 (0%) | 0 (0%) | 43 (4.0%) | |
| Total cholesterol | ||||
| Mean ± SD | 225 ± 43.1 | 224 ± 44.4 | 213 ± 45.9 | <0.001 |
| Missing | 28 (6.5%) | 71 (9.0%) | 2 (0.2%) | |
| HDL cholesterol | ||||
| Mean ± SD | 51.9 ± 14.5 | 52.3 ± 15.6 | 53.2 ± 14.2 | 0.198 |
| Missing | 30 (7.0%) | 38 (4.8%) | 1 (0.1%) | |
| Triglycerides | ||||
| Mean ± SD | 114 ± 74.4 | 114 ± 70.7 | 117 ± 76.0 | 0.583 |
| Missing | 28 (6.5%) | 34 (4.3%) | 1 (0.1%) | |
| LDL cholesterol | ||||
| Mean ± SD | 152 ± 38.4 | 149 ± 38.6 | 136 ± 39.7 | <0.001 |
| Missing | 43 (10.0%) | 98 (12.5%) | 27 (2.5%) | |
| History of hyperchol. | ||||
| Yes | 97 (22.5%) | 256 (32.6%) | 356 (33.1%) | <0.001 |
| No | 334 (77.5%) | 515 (65.5%) | 715 (66.4%) | |
| Missing | 0 (0%) | 15 (1.9%) | 6 (0.6%) | |
| Cholesterol treatment | ||||
| No | 403 (93.5%) | 705 (89.7%) | 903 (83.8%) | <0.001 |
| Yes | 28 (6.5%) | 68 (8.7%) | 132 (12.3%) | |
| Missing | 0 (0%) | 13 (1.7%) | 42 (3.9%) | |
| Height (cm) | ||||
| Mean ± SD | 163 ± 9.21 | 162 ± 9.39 | 163 ± 9.05 | 0.00368 |
| Missing | 8 (1.9%) | 15 (1.9%) | 12 (1.1%) | |
| Weight (Kg) | ||||
| Mean ± SD | 72.3 ± 12.6 | 73.8 ± 14.0 | 73.6 ± 13.9 | 0.12 |
| Missing | 8 (1.9%) | 15 (1.9%) | 12 (1.1%) | |
| Body mass index | ||||
| Mean ± SD | 27.0 ± 4.15 | 28.1 ± 4.62 | 27.6 ± 4.63 | <0.001 |
| Missing | 8 (1.9%) | 15 (1.9%) | 12 (1.1%) | |
| Physical activity (Kcal/week) | ||||
| Mean ± SD | 491 ± 419 | 422 ± 377 | 351 ± 378 | <0.001 |
| Missing | 64 (14.8%) | 22 (2.8%) | 2 (0.2%) | |
| Physical component | ||||
| Mean ± SD | 49.3 ± 8.08 | 49.0 ± 9.63 | 50.1 ± 8.91 | 0.0366 |
| Missing | 34 (7.9%) | 123 (15.6%) | 83 (7.7%) | |
| Mental component | ||||
| Mean ± SD | 49.2 ± 11.3 | 48.9 ± 11.0 | 46.9 ± 10.8 | <0.001 |
| Missing | 34 (7.9%) | 123 (15.6%) | 83 (7.7%) | |
| Cardiovascular event | ||||
| No | 388 (90.0%) | 706 (89.8%) | 977 (90.7%) | 0.161 |
| Yes | 10 (2.3%) | 35 (4.5%) | 47 (4.4%) | |
| Missing | 33 (7.7%) | 45 (5.7%) | 53 (4.9%) | |
| Days to cardiovascular event or end of follow-up | ||||
| Mean ± SD | 1780 ± 1100 | 1690 ± 1080 | 1790 ± 1070 | 0.0989 |
| Missing | 33 (7.7%) | 45 (5.7%) | 53 (4.9%) | |
| Overall death | ||||
| No | 369 (85.6%) | 657 (83.6%) | 949 (88.1%) | <0.001 |
| Yes | 18 (4.2%) | 81 (10.3%) | 74 (6.9%) | |
| Missing | 44 (10.2%) | 48 (6.1%) | 54 (5.0%) | |
| Days to overall death or end of follow-up | ||||
| Mean ± SD | 1710 ± 1040 | 1670 ± 1050 | 1760 ± 1050 | 0.254 |
| Missing | 44 (10.2%) | 48 (6.1%) | 54 (5.0%) |
47.5 和CompareGroups比较
CompareGroups不能自定义复杂表头。但是胜在语法简单，自带P值！详情请参考前面的文章。