R ifelse 多个条件

Question

我的数据就像

用户名	复合分数	年龄	日期
一个	0.5	26	2020-08-10
一个	0.6	26	2020-09-01
乙	0.3	27	2020-11-15

structure(list(age = c(24L, 28L, 25L, 27L, 30L, 25L, 47L, 23L, 
26L, 23L), compound = c(-0.765, 0.743, 0.1901, 0, 0, 0.743, 0.2732, 
-0.2263, 0.3612, -0.2263), Date = structure(c(18551, 18544, 18544, 
18541, 18540, 18538, 18536, 18536, 18534, 18533), class = "Date")), row.names = c(1L, 
2L, 3L, 4L, 5L, 6L, 7L, 9L, 10L, 12L), class = "data.frame")

我拥有的代码将数据转换为具有两个条件的长格式。

> twitter_wide = twitter_d %>% mutate(Date = as.Date(Date)) %>%
+   group_by(username,age,Trial = ifelse(day(Date) >= 3 & month(Date) >11, 2, 1)) %>%
+   summarise(compound = mean(compound), .groups = 'drop')

如何将数据更改为三个条件和宽格式：

条件1：平均分数（复合）日期从8.16-11.2（T1）

条件 2：平均分数（复合）日期从 11.3-1.20 (T2)

条件3：平均分数（复合）日期从1.21-4.9（T3）

所需的表可能看起来像

用户名	年龄	Mean_compound score_T1	Mean_compound score_T2	Mean_compound score_T3
一个	26	0.5	0.3	0.7
乙	26	0.3	0.5	0.3

Answer 1

这是一个假数据的例子：

# example data
twitter_d <- structure(list(username = c("A", "A", "A", "A", "B", "B", "B", "B"), 
    `compound score` = c(0.5, 0.6, 0.7, 0.8, 0.3, 0.2, 0.1, 0), 
    age = c(26L, 26L, 26L, 26L, 27L, 27L, 27L, 27L), 
    Date = c("2020-08-16", "2020-09-01", "2020-11-14", "2021-01-20", "2020-09-12", "2020-11-02", "2020-11-15", "2021-04-09")),
        class = "data.frame", row.names = c(NA, -8L))

# solution
library(tidyverse)

twitter_wide <- twitter_d %>% mutate(
    Condition = ifelse(Date >= "2020-08-16" & Date <= "2020-11-02", "T1", 
                    ifelse(Date >= "2020-11-03" & Date <= "2021-01-20", "T2", 
                        ifelse(Date >= "2021-01-21" & Date <= "2021-04-09", "T3", NA)))
) %>% 
    group_by(username, age, Condition) %>% 
    summarise(compound = mean(`compound score`, na.rm = TRUE), .groups = "drop") %>% 
    pivot_wider(names_from = Condition, names_prefix = "Mean_compound_score_", values_from = compound)

twitter_wide
## A tibble: 2 x 5
#  username   age Mean_compound_score_T1 Mean_compound_score_T2 Mean_compound_score_T3
#  <chr>    <int>                  <dbl>                  <dbl>                  <dbl>
#1 A           26                   0.55                   0.75                     NA
#2 B           27                   0.25                   0.1                       0

解释：

我mutate()一个名为Condition的新列。 它包含c("T1", "T2", "T3")取决于为每个时间段指定的Date范围。 基本上，这是对日期进行分组。

group_by()和summarise()表示您想要的 function （您已经拥有）。

pivot_wider()将数据更改为宽格式，使用来自compound的值并按日期分组列Condition对它们进行分组。

有关来自tidyr package 的更多信息和用于pivot_wider()的信息，请参阅https://tidyr.tidyverse.org/reference/pivot_wider.ZFC35FDC70D5FC69D53EZ8C 。