问题描述
我有这个带有可重现数据的 df：
for z in web_elements:
print(z.get_attribute("innerHTML"))
width = browser.find_element_by_xpath('//input[contains(@class,"cke_dialog_ui_input_text")]')
width.click()
并且我通过使用以下代码旋转 df 创建了一个数据框来计算分类观察：
structure(list(`Loperamida en diarrea` = structure(c(1L,1L,2L,4L,3L,1L),.Label = c("muy efectiva","algo efectiva","no efectiva","no se"),class = c("ordered","factor")),`Carbón en diarrea` = structure(c(2L,4L),`Bismuto en diarrea` = structure(c(2L,`Rifaximina en diarrea` = structure(c(2L,`Otros antibióticos en diarrea` = structure(c(2L,`Probióticos en diarrea` = structure(c(2L,2L),`Orientación dicotómica` = c("Neurogastro","Neurogastro","No neurogastro","No neurogastro")),row.names = c(NA,10L),class = "data.frame")
结果如下：
library(tidyverse)
library(janitor)
df %>%
pivot_longer(cols = everything()) %>%
count(name,value) %>%
pivot_wider(names_from = value,values_from = n,values_fill = 0) %>%
mutate("efectiva" = `algo efectiva` + `muy efectiva`) %>%
arrange(desc(`efectiva`)) %>%
select(c(`name`,`efectiva`,`no efectiva`)) %>%
adorn_percentages("row") %>%
adorn_pct_formatting(digits = 1) %>%
adorn_ns()
df$name <- str_remove(df$name," en diarrea")
我一直在尝试通过变量 name efectiva no efectiva
Rifaximina 100.0% (10) 0.0% (0)
Probióticos 80.0% (8) 20.0% (2)
Bismuto 77.8% (7) 22.2% (2)
Loperamida 87.5% (7) 12.5% (1)
Otros antibióticos 77.8% (7) 22.2% (2)
Carbón 50.0% (3) 50.0% (3)
Orientación dicotómica - (0) - (0)
（Neurogastro vs No Neurogastro）来分隔列，但我一直无法解决。我期望的是这样的：
Orientación dicotómica有什么建议吗？
解决方法
编辑虽然有点延迟
library(janitor)
library(tidyverse)
df %>%
pivot_longer(cols = 1:6) %>%
count(`Orientación dicotómica`,name,value) %>%
pivot_wider(id_cols = c(`Orientación dicotómica`,name),names_from = value,values_from = n,values_fill = 0,values_fn = sum) %>%
mutate("efectiva" = `algo efectiva` + `muy efectiva`) %>%
select(c(`Orientación dicotómica`,`name`,`efectiva`,`no efectiva`)) %>%
adorn_percentages("row") %>%
adorn_pct_formatting(digits = 1) %>%
adorn_ns() -> out
merge(out %>% filter(`Orientación dicotómica` == 'Neurogastro') %>% select(name,`Neurogastro efectiva` = efectiva,`Neurogastro no efectiva` = `no efectiva`),out %>% filter(`Orientación dicotómica` == 'No neurogastro') %>% select(name,`No Neurogastro efectiva` = efectiva,`No Neurogastro no efectiva` = `no efectiva`),by = "name")
name Neurogastro efectiva Neurogastro no efectiva No Neurogastro efectiva No Neurogastro no efectiva
1 Bismuto en diarrea 66.7% (4) 33.3% (2) 100.0% (3) 0.0% (0)
2 Carbón en diarrea 75.0% (3) 25.0% (1) 0.0% (0) 100.0% (2)
3 Loperamida en diarrea 75.0% (3) 25.0% (1) 100.0% (4) 0.0% (0)
4 Otros antibióticos en diarrea 100.0% (6) 0.0% (0) 33.3% (1) 66.7% (2)
5 Probióticos en diarrea 100.0% (6) 0.0% (0) 50.0% (2) 50.0% (2)
6 Rifaximina en diarrea 100.0% (6) 0.0% (0) 100.0% (4) 0.0% (0)
这不是我想要的，但非常接近：
## fist I create 2 df's with filter:
library(dplyr)
df1 <- df %>% filter(`Orientación dicotómica` == "Neurogastro")
df2 <- df %>% filter(`Orientación dicotómica` != "Neurogastro")
## then I bind the 2 df's
df3 <- cbind(df1,df2)
## finally I drop the repeated column and create a new df with renamed columns
df4 <- as.data.frame(as.matrix(df3[-4]) %>%
list(name = df_tto$name,Neurogastro = df3[,c(2,3)],No_neurogastro = df3[,c(5,6)]))
df4 <- df4[-(16),-(2:6)]
对此可能有更好的答案以及简化的代码，但这就是我能想到的全部内容，无论如何，它几乎可以完成工作......