geom_segment y-轴的重排顺序
geom_segment y-轴的重排顺序
提问于 2021-02-12 21:35:15
问题：想要按geom_segment的长度来排序ggplot。
数据
swimplot_response
SUBJECT RESPONSE ORD_RESP STARTRESP ENDRESP ORD DEATH MAX_RESP
1 01 On Study 1 0.0000000 88.9000000 1 1 94.483871
2 01 Stable Disease 2 0.0000000 9.2333333 2 1 94.483871
3 01 Partial Response 3 9.2333333 21.0333333 3 1 94.483871
4 01 Complete Response 4 21.0333333 62.0333333 4 1 94.483871
5 01 Progressive Disease -1 62.0333333 88.0000000 5 1 94.483871
6 01 Off Study/Alive 0 88.0000000 94.4838710 6 1 94.483871
7 02 On Study 1 0.0000000 93.5806452 1 1 93.806452
8 02 Stable Disease 2 0.0000000 1.9000000 2 1 93.806452
9 02 Progressive Disease -1 1.9000000 93.5806452 3 1 93.806452
10 02 Off Study/Alive 0 93.5806452 93.8064516 4 1 93.806452
11 03 On Study 1 0.0000000 4.1290323 1 1 5.774194
12 03 Stable Disease 2 0.0000000 0.6666667 2 1 5.774194
13 03 Progressive Disease -1 0.6666667 4.1290323 3 1 5.774194
14 03 Off Study/Alive 0 4.1290323 5.7741935 4 1 5.774194
15 04 On Study 1 0.0000000 2.5666667 1 1 5.607143
16 04 Stable Disease 2 0.0000000 2.5666667 2 1 5.607143
17 04 Progressive Disease -1 2.5666667 5.6071429 3 1 5.607143
18 05 On Study 1 0.0000000 11.0645161 1 1 19.833333
19 05 Stable Disease 2 0.0000000 3.0000000 2 1 19.833333
20 05 Progressive Disease -1 3.0000000 11.0645161 3 1 19.833333
21 05 Off Study/Alive 0 11.0645161 19.8333333 4 1 19.833333，
- ，然后我用
MAX_RESP来细分。在这段代码中，我认为我已经解决了由MAX_RESP.
安排dfs的问题。
subj_18mos_resp <- swimplot_response %>%
group_by(SUBJECT) %>%
filter(MAX_RESP >= 18) %>%
arrange(desc(MAX_RESP))
subj_18mos_resp
subj_18mos <- unique(subj_18mos_resp$SUBJECT)
subj_18mos_bev <- swimplot_bev %>%
subset(SUBJECT %in% subj_18mos) %>%
group_by(SUBJECT) %>%
arrange(ORD, .by_group = TRUE)
subj_18mos_steroids <- swimplot_steroids %>%
subset(SUBJECT %in% subj_18mos) %>%
group_by(SUBJECT) %>%
arrange(ORD, .by_group = TRUE)
subj_18mos_dose <- swimplot_dose %>%
subset(SUBJECT %in% subj_18mos) %>%
group_by(SUBJECT)
subj_18mos_resp
# A tibble: 66 x 8
# Groups: SUBJECT [16]
SUBJECT RESPONSE ORD_RESP STARTRESP ENDRESP ORD DEATH MAX_RESP
<chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 01 On Study 1 0 88.9 1 1 94.5
2 01 Stable Disease 2 0 9.23 2 1 94.5
3 01 Partial Response 3 9.23 21.0 3 1 94.5
4 01 Complete Response 4 21.0 62.0 4 1 94.5
5 01 Progressive Disease -1 62.0 88 5 1 94.5
6 01 Off Study/Alive 0 88 94.5 6 1 94.5
7 02 On Study 1 0 93.6 1 1 93.8
8 02 Stable Disease 2 0 1.9 2 1 93.8
9 02 Progressive Disease -1 1.9 93.6 3 1 93.8
10 02 Off Study/Alive 0 93.6 93.8 4 1 93.8
# ... with 56 more rows- I还尝试按当前顺序将
SUBJECT转换为因子和级别，但这没有起作用。(错误消息包含在下面。)
subj_18mos_resp$SUBJECT <- factor(levels = subj_18mos_resp$SUBJECT)
Error in `levels<-`(`*tmp*`, value = as.character(levels)) :
factor level [2] is duplicated我也尝试过在ggplot(aes())中包括subj_18mos_resp
我的绘图代码包括在下面，以及示例图的图像。
“全”游泳图代码(含剂量，bev，类固醇使用)，适用于总生存期大于18 dose的受试者
SwimmerPlot_18mos_full <- ggplot() +
theme_classic() +
geom_vline(xintercept = 6, colour = "black", linetype = "longdash",
show.legend = FALSE) +
geom_segment(data = subj_18mos_resp, aes(x = STARTRESP, xend = ENDRESP,
y = SUBJECT, yend = SUBJECT,
colour = RESPONSE),
size = 4, show.legend = TRUE) +
scale_colour_manual(values = response_color) +
geom_segment(data = subj_18mos_steroids, aes(x = STARTROIDS, xend = ENDROIDS,
y = SUBJECT, yend = SUBJECT,
colour = STEROIDS),
size = 2, linetype = 1, show.legend = TRUE) +
geom_segment(data = subj_18mos_bev, aes(x = STARTBEV, xend = ENDBEV,
y = SUBJECT, yend = SUBJECT,
colour = BEV),
size = 1, linetype = 1, show.legend = TRUE) +
geom_point(data = subj_18mos_dose, aes(x = DUR, y = SUBJECT,
fill = TREATMENT),
shape = 25, size = 2.5, show.legend = FALSE) +
labs(title = "Phase 1 GBM Swim Plots",
subtitle = "Data from study [redacted]",
caption = "Subjects with OS of >18 mos.",
color = "Response") +
scale_x_continuous(name = "Duration in Months", breaks = seq(0,90,10)) +
ylab(label = "Subject ID") +
scale_fill_discrete() +
geom_vline(xintercept = 6, colour = "black", linetype = "longdash",
show.legend = FALSE) +
geom_text(aes(x=DUR, y=SUBJECT, label = TREATMENT), data = subj_18mos_dose,
nudge_y = 0.5, nudge_x = 0.5, size = 2.5) +
guides(fill = FALSE)
SwimmerPlot_18mos_full
受试者>18莫斯反应
SwimmerPlot_18mos <- ggplot() +
geom_segment(data = subj_18mos_resp, aes(x = STARTRESP, xend = ENDRESP,
y = SUBJECT, yend = SUBJECT,
colour = RESPONSE),
size = 4, show.legend = TRUE) +
scale_colour_manual(values = response_color) +
labs(title = "Phase 1 GBM Swim Plots",
subtitle = "Data from study [redacted]",
caption = "Subjects with OS of >18 mos.",
color = "Response") +
scale_x_continuous(name = "Duration in Months", breaks = seq(0,90,10)) +
ylab(label = "Subject ID") +
#Add dashed line at 6 month mark
geom_vline(xintercept = 6, colour = "black", linetype = "longdash",
show.legend = TRUE) +
theme_classic()
SwimmerPlot_18mos回答 1
Stack Overflow用户
回答已采纳
发布于 2021-02-15 16:01:58
非常感谢格雷戈·托马斯！
最后，我通过升序和重构来安排MAX_RESP，因为ggplot默认按升序构建级别。您的评论将我引向以下解决方案(标记为#comments的更改)：
subj_18mos_resp <- swimplot_response %>%
group_by(SUBJECT) %>%
filter(MAX_RESP >= 18) %>%
arrange(MAX_RESP) #Changed from arrange(desc(MAX_RESP))
subj_18mos_resp
subj_18mos <- unique(subj_18mos_resp$SUBJECT)
subj_18mos_resp$SUBJECT <- factor(subj_18mos_resp$SUBJECT, levels = subj_18mos)
#Re-specified levels from the unique vector produced when arranged by increasing MAX_RESP
subj_18mos_bev <- swimplot_bev %>%
subset(SUBJECT %in% subj_18mos) %>%
group_by(SUBJECT) %>%
arrange(ORD, .by_group = TRUE)
subj_18mos_steroids <- swimplot_steroids %>%
subset(SUBJECT %in% subj_18mos) %>%
group_by(SUBJECT) %>%
arrange(ORD, .by_group = TRUE)
subj_18mos_dose <- swimplot_dose %>%
subset(SUBJECT %in% subj_18mos) %>%
group_by(SUBJECT)
页面原文内容由Stack Overflow提供。腾讯云小微IT领域专用引擎提供翻译支持
原文链接：
https://stackoverflow.com/questions/66179334
复制相关文章
相似问题
腾讯云开发者
Copyright © 2013 - 2026 Tencent Cloud. All Rights Reserved. 腾讯云 版权所有
深圳市腾讯计算机系统有限公司 ICP备案/许可证号：粤B2-20090059 
粤公网安备44030502008569号
腾讯云计算（北京）有限责任公司 京ICP证150476号 | 京ICP备11018762号