# df %>%
# mutate(new_var = case_when(var1 < 15 ~ 'low',
# var2 < 25 ~ 'med',
# TRUE ~ 'high'))
df <- data.frame(player = c('AJ', 'Bob', 'Chad', 'Dan', 'Eric', 'Frank'),
position = c('G', 'F', 'F', 'G', 'C', NA),
points = c(12, 15, 19, 22, 32, NA),
assists = c(5, 7, 7, 12, 11, NA))
# player position points assists
# 1 AJ G 12 5
# 2 Bob F 15 7
# 3 Chad F 19 7
# 4 Dan G 22 12
# 5 Eric C 32 11
# 6 Frank
NA NA
注：TRUE相当于else条件语句。首先我们准备示例数据，为后面示例使用。
下面代码展示如何创建新的变量quality，其值有point变量产生：
df %>%
mutate(quality = case_when(points > 20 ~ 'high',
points > 15 ~ 'med',
TRUE ~ 'low' ))
# player position points assists quality
# 1 AJ G 12 5 low
# 2 Bob F 15 7 low
# 3 Chad F 19 7 med
# 4 Dan G 22 12 high
# 5 Eric C 32 11 high
# 6 Frank <NA> NA NA low
上面代码比较好理解：point>20 则 quality 为 high, point>15 则 quality 为 med,其他情况(包括NA)为low。
下面示例展示基于 points 和 assists 两个变量产生quality变量：
df %>%
mutate(quality = case_when(points > 15 & assists > 10 ~ 'great',
points > 15 & assists > 5 ~ 'good',
TRUE ~ 'average' ))
# player position points assists quality
# 1 AJ G 12 5 average
# 2 Bob F 15 7 average
# 3 Chad F 19 7 good
# 4 Dan G 22 12 great
# 5 Eric C 32 11 great
# 6 Frank <NA> NA NA average
df %>%
mutate(quality = case_when(is.na(points) ~ 'missing',
points > 15 & assists > 10 ~ 'great',
points > 15 & assists > 5 ~ 'good',
TRUE ~ 'average' ))
# player position points assists quality
# 1 AJ G 12 5 average
# 2 Bob F 15 7 average
# 3 Chad F 19 7 good
# 4 Dan G 22 12 great
# 5 Eric C 32 11 great
# 6 Frank <NA> NA NA missing
后面代码使用is.na对NA记录单独进行判断。另外需要说明的是case_when 中的几个条件语句顺序很重要，TRUE语句一定要放在最后。
每个学生有三个学科成绩，但分值范围不同。我们首先要对每个学科成绩标准化，然后再对每个学生成绩求平均值，最后根据设定范围生成等级。
library(dplyr)
options(digits=2)
Student <- c("John Davis", "Angela Williams", "Bullwinkle Moose",
"David Jones", "Janice Markhammer", "Cheryl Cushing",
"Reuven Ytzrhak", "Greg Knox", "Joel England",
"Mary Rayburn")
Math <- c(502, 600, 412, 358, 495, 512, 410, 625, 573, 522)
Science <- c(95, 99, 80, 82, 75, 85, 80, 95, 89, 86)
English <- c(25, 22, 18, 15, 20, 28, 15, 30, 27, 18)
roster <- data.frame(Student, Math, Science, English, stringsAsFactors=FALSE)
z <- roster %>% mutate_if(is.numeric, scale)
# roster %>% select(-Student) %>% mutate_each(funs(scale))
z <- z %>% mutate(score = rowMeans(select(.,-Student)))
y <- quantile(score, c(0.8, 0.6, 0.4, 0.2))
# Student Math Science English score
# 1 John Davis 0.013 1.078 0.587 0.56
# 2 Angela Williams 1.143 1.591 0.037 0.92
# 3 Bullwinkle Moose -1.026 -0.847 -0.697 -0.86
# 4 David Jones -1.649 -0.590 -1.247 -1.16
# 5 Janice Markhammer -0.068 -1.489 -0.330 -0.63
# 6 Cheryl Cushing 0.128 -0.205 1.137 0.35
# 7 Reuven Ytzrhak -1.049 -0.847 -1.247 -1.05
# 8 Greg Knox 1.432 1.078 1.504 1.34
# 9 Joel England 0.832 0.308 0.954 0.70
# 10 Mary Rayburn 0.243 -0.077 -0.697 -0.18
z <- z %>% mutate(
level = case_when(
score >= y[1] ~ 'A',
score < y[1] & score >= y[2] ~ 'B',
score < y[2] & score >= y[3] ~ 'C',
score < y[3] & score >= y[4] ~ 'D',
score < y[4] ~ 'F'
# Student Math Science English score level
# 1 John Davis 0.013 1.078 0.587 0.56 B
# 2 Angela Williams 1.143 1.591 0.037 0.92 A
# 3 Bullwinkle Moose -1.026 -0.847 -0.697 -0.86 D
# 4 David Jones -1.649 -0.590 -1.247 -1.16 F
# 5 Janice Markhammer -0.068 -1.489 -0.330 -0.63 D
# 6 Cheryl Cushing 0.128 -0.205 1.137 0.35 C
# 7 Reuven Ytzrhak -1.049 -0.847 -1.247 -1.05 F
# 8 Greg Knox 1.432 1.078 1.504 1.34 A
# 9 Joel England 0.832 0.308 0.954 0.70 B
# 10 Mary Rayburn 0.243 -0.077 -0.697 -0.18 C
代码连接起来，一气呵成：
roster %>% mutate_if(is.numeric, scale) %>%
mutate(score = rowMeans(select(.,-Student))) %>%
mutate(
level = case_when(
score >= y[1] ~ 'A',
score < y[1] & score >= y[2] ~ 'B',
score < y[2] & score >= y[3] ~ 'C',
score < y[3] & score >= y[4] ~ 'D',
score < y[4] ~ 'F'
dplyr提供的case_when()函数简化多个if_else语句，如果不匹配返回NA。结合mutate函数可基于已有变量创建新的变量。本文通过几个示例进行对比学习。语法：library(dplyr)# df %>%# mutate(new_var = case_when(var1 < 15 ~ 'low',# var2 < 25 ~ 'med',# .
1、case_when函数，有一些SQL基础（casewhen）的都猜得到这个函数的功能
可实现多条件判断并可以添加标签的函数，这在我们对数据进行分类整理中十分的实用，这个函数中的参数可以这样分：一部分是判断条件，另一部分是所要做的标签
iris%>%select(Sepal.Lengt...
R包dplyr可用于处理R内部或者外部的结构化数据，相较于plyr包，dplyr专注接受dataframe对象, 大幅提高了速度,
并且提供了更稳健的数据库接口。同时，dplyr包可用于操作Spark的dataframe。
1. 数据集类型转换
tbl_df()可用于将过长过大的数据集转换为显示更友好的 tbl_df 类型。使用dplyr包处理数据前，建议先将数据集转换为tbl对象。
1. CASE WHEN 表达式有两种形式
--简单Case函数 CASE sex WHEN '1' THEN '男' WHEN '2' THEN '女' ELSE '其他' END --Case搜索函数 CASEWHEN sex = '1' THEN '男' WHEN sex = '2' THEN '女' ELSE '其他' END
2. CASE WHEN...
