英文:
How to mask certain cell's info from a gtsummary table
问题
我有一些敏感信息在我的数据中，我需要将其隐藏在某个特定的阈值以下（以符合数据使用协议并防止重新识别数据）。我正在使用gtsummary中的tbl_svysmmary()。在我的示例中，我想要过滤"cell size ≤ 100"：
library(gtsummary)library(survey)tbl_svysummary <-svydesign(~1, data = as.data.frame(Titanic), weights = ~Freq) %>%tbl_svysummary(by = Survived, percent = "row", include = c(Class, Age))tbl_svysummary
我想显示儿童信息如下：
编辑：用于多个stat_列和0/1变量的可重现示例：
library(gtsummary)library(survey)supp_outcomes <-as.data.frame(Titanic) %>%mutate(Female = ifelse(Sex == "Female", 1, 0)) %>%svydesign(~1, data = ., weights = ~Freq) %>%tbl_svysummary(by = Survived, percent = "row",include = c(Age, Female, Class)) %>%add_overall() %>%add_p()supp_outcomes
从@Marco建议的解决方案中编辑的代码：
supp_outcomes$table_body <- supp_outcomes$table_body %>%mutate(extra1 = stat_0,extra2 = stat_1,extra3 = stat_2) %>%separate(extra1, c("number"), sep = ' \\(') %>%separate(extra2, c("number"), sep = ' \\(') %>%separate(extra3, c("number"), sep = ' \\(') %>%mutate(number = as.numeric(number)) %>%mutate(stat_0 = case_when(number < 200 & number > 0 & var_type %in% c("dichotomous", "categorical") ~ "TOO FEW",TRUE ~ stat_0),stat_1 = case_when(number < 200 & number > 0 & var_type %in% c("dichotomous", "categorical") ~ "TOO FEW",TRUE ~ stat_1),stat_2 = case_when(number < 200 & number > 0 & var_type %in% c("dichotomous", "categorical") ~ "TOO FEW",TRUE ~ stat_2)) %>%select(!number)supp_outcomes
英文:
I have some sensitive info in my data that I need to hide below a certain threshold (to comply with DUA and prevent reidentifying data). I'm using tbl_svysmmary() from gtsummary. In my example, I like to filter "cell size ≤ 100":
library(gtsummary)library(survey)tbl_svysummary <-svydesign(~1, data = as.data.frame(Titanic), weights = ~Freq) %>%tbl_svysummary(by = Survived, percent = "row", include = c(Class, Age))tbl_svysummary
I want to show child info as:
EDIT: Reproducible example for multiple stat_ columns and 0/1 variables:
library(gtsummary)library(survey)supp_outcomes <-as.data.frame(Titanic) %>%mutate (Female=ifelse(Sex=="Female",1,0)) %>%svydesign(~1, data = ., weights = ~Freq) %>%tbl_svysummary(by = Survived, percent = "row",include = c(Age, Female, Class)) %>%add_overall() %>% add_p()supp_outcomes
The edited code from @Marco's suggested solution:
supp_outcomes$table_body <- supp_outcomes$table_body %>%mutate(extra1 = stat_0,extra2 = stat_1,extra3 = stat_2) %>%separate(extra1, c("number"), sep = ' \\(') %>%separate(extra2, c("number"), sep = ' \\(') %>%separate( extra3, c("number"), sep = ' \\(') %>%mutate(number = as.numeric(number)) %>%mutate(stat_0 = case_when(number < 200 & number > 0 & var_type %in%c("dichotomous", "categorical")~ "TOO FEW",TRUE ~ stat_0),stat_1 = case_when(number < 200 & number > 0 & var_type %in%c("dichotomous", "categorical")~ "TOO FEW",TRUE ~ stat_1),stat_2 = case_when(number < 200 & number > 0 & var_type %in%c("dichotomous", "categorical")~ "TOO FEW",TRUE ~ stat_2)) %>%select(!number)supp_outcomes
答案1
得分: 1
你可以对输出对象进行更多的 tidyverse 操作，就像这样（使用 table_body）：
library(tidyverse)data(mtcars)library(gtsummary)output <- mtcars[,1:2] %>% tbl_summary()output$table_body# 一个 tibble: 5 × 6variable var_type var_label row_type label stat_0<chr> <chr> <chr> <chr> <chr> <chr>1 mpg continuous mpg label mpg 19.2 (15.4, 22.8)2 cyl categorical cyl label cyl NA3 cyl categorical cyl level 4 11 (34%)4 cyl categorical cyl level 6 7 (22%)5 cyl categorical cyl level 8 14 (44%)# 当 cyl 的 N 少于 10 时，不显示单元格信息output$table_body <- output$table_body %>%mutate(extra = stat_0) %>%separate(extra, c("number"), sep = ' \\(') %>%mutate(number = as.numeric(number)) %>%mutate(stat_0 = case_when(number < 10 & var_type == "categorical" ~ "TOO FEW",TRUE ~ stat_0)) %>%select(!number)output
case_when 中的筛选决策基于分类变量，当单元格信息少于阈值时，你可以控制它。你可以根据其他变量或条件进行调整。
英文:
You can do more tidyverse manipulation on the output object like this (using the table_body):
library(tidyverse)data(mtcars)library(gtsummary)output <- mtcars[,1:2] %>% tbl_summary()output$table_body# A tibble: 5 × 6variable var_type var_label row_type label stat_0<chr> <chr> <chr> <chr> <chr> <chr>1 mpg continuous mpg label mpg 19.2 (15.4, 22.8)2 cyl categorical cyl label cyl NA3 cyl categorical cyl level 4 11 (34%)4 cyl categorical cyl level 6 7 (22%)5 cyl categorical cyl level 8 14 (44%)# Don't show cell information when N of cyl is less than 10output$table_body <- output$table_body %>%mutate(extra = stat_0) %>%separate(extra, c("number"), sep = ' \\(') %>%mutate(number = as.numeric(number)) %>%mutate(stat_0 = case_when(number < 10 & var_type == "categorical" ~ "TOO FEW",TRUE ~ stat_0)) %>%select(!number)output
The filter decision in case_when is based on categorical variables, where you like to control the cell information, when it is less than a threshold value. You can adjust this for any other variable or condition.
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论