R语言：转换—宽数据↔长数据

博客来源于我的语雀专栏：R 语言 · 语雀

更多内容同步更新请关注我的语雀：令平子 · 语雀

参考资料：部分已引用到各段落

宽 转 长

二维数据 转 长

r - Reshaping data.frame from wide to long format - Stack Overflow

#create data frame 
df <- data.frame(index=c(1, 2, 3, 4, 5, 6),
                 group=c('a','b','c','d','e','f'),
                 var1=c(4, 4, 5, 4, 3, 2),
                 var2=c(1, 2, 4, 4, 6, 9),
                 var3=c(9, 9, 9, 5, 5, 3))

方法一：reshape2::melt

long <- reshape2::melt(df , id.vars = c('index','group'), variable.name = 'series')

# index group series value

# 1 1 a var1 4

# 2 2 b var1 4

# 3 3 c var1 5

其他等效代码：

melt(df, id.vars = 1:2, variable.name = "series")

melt(df, measure.vars = 3:5, variable.name = "series")

melt(df, measure.vars = c("var1","var2","var3"), variable.name = "series")

方法二：tidyr::pivot_longer

long <- tidyr::pivot_longer(df,cols = c("var1","var2","var3"), names_to = "series",values_to = "value")

long <- tidyr::pivot_longer(df,cols =-c('index','group'), names_to = "series",values_to = "value")

方法三：reshape

long <- reshape(df, direction = "long",varying = c("var1","var2","var3"),v.names = "value",idvar = c('index','group'))

# index group time value

# 1: 1 a 1 4

# 2: 2 b 1 4

# 3: 3 c 1 5

一维数据框 转 长

r - Convert data.frame column format from character to factor - Stack Overflow

一维数据框转long的核心就是将分类列设置成因子类型

df <- data.frame(index=c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20),
                 series=c('a','b','c','d','a','b','c','d','a','b','c','d','a','b','c','d','a','b','c','d'),
                 value=c(94,79,24,87,7,93,99,23,26,32,96,27,42,5,70,16,98,91,21,55))
#     id series value
# 1   1    a    94
# 2   2    b    79
# 3   3    c    24
# 4   4    d    87
# 5   5    a     7
# 6   6    b    93
# 7   7    c    99
# 8   8    d    23
# 9   9    a    26
# 10 10    b    32
# 11 11    c    96
# 12 12    d    27
# 13 13    a    42
# 14 14    b     5
# 15 15    c    70
# 16 16    d    16
# 17 17    a    98
# 18 18    b    91
# 19 19    c    21
# 20 20    d    55

方法一：melt

library(data.table)

long <- melt(df, id.vars = c("id"), variable.name = "series")

方法二：reshape

long <- reshape(

df,

idvar = "series", # 对象标识符

v.names = "value", # 观测值

direction = "long") # 转换为长格式

方法三：单列、多列设置为factor 推荐

df$series<- as.factor(df$series) 或 df[, 'series'] <- as.factor(df[, 'series'])

多列设置

col_names <- c("series")

df[col_names] <- lapply(df[col_names] , factor)

长 转 宽

r - Convert data from long format to wide format with multiple measure columns - Stack Overflow