如何根据多个条件对行求和 - R？ [复制]

Posted 2023-02-19

【中文标题】如何根据多个条件对行求和 - R？ [复制]

【英文标题】：How to sum rows based on multiple conditions - R? [duplicate]

【发布时间】：2015-05-09 11:56:05

【问题描述】：

我有一个数据框，其中包含地块 ID (plotID)、树种代码 (species) 和覆盖值 (cover)。您可以看到其中一个地块中有多个树种记录。如果每个地块中有重复的“物种”行，我如何对“封面”字段求和？

例如，这里是一些示例数据：

# Sample Data
plotID = c( "SUF200001035014", "SUF200001035014", "SUF200001035014", "SUF200001035014", "SUF200001035014", "SUF200046012040",
       "SUF200046012040", "SUF200046012040", "SUF200046012040", "SUF200046012040", "SUF200046012040", "SUF200046012040")
species = c("ABBA",  "BEPA",  "PIBA2", "PIMA",  "PIRE",  "PIBA2", "PIBA2", "PIMA",  "PIMA",  "PIRE",  "POTR5", "POTR5")
cover = c(26.893939,  5.681818,  9.469697, 16.287879,  1.893939, 16.287879,  4.166667, 10.984848, 16.666667, 11.363636, 18.181818,
          13.257576)
df_original = data.frame(plotID, species, cover)

这是预期的输出：

# Intended Output
plotID2 = c( "SUF200001035014", "SUF200001035014", "SUF200001035014", "SUF200001035014", "SUF200001035014", "SUF200046012040",
            "SUF200046012040", "SUF200046012040", "SUF200046012040")
species2 = c("ABBA",  "BEPA",  "PIBA2", "PIMA",  "PIRE",  "PIBA2", "PIMA",  "PIRE",  "POTR5")
cover2 = c(26.893939,  5.681818,  9.469697, 16.287879,  1.893939, 20.454546, 18.651515, 11.363636, 31.439394)
df_intended_output = data.frame(plotID2, species2, cover2)

【参考方案1】：

aggregate 很容易

aggregate(cover~species+plotID, data=df_original, FUN=sum) 

data.table 更容易

as.data.table(df_original)[, sum(cover), by = .(plotID, species)]

【参考方案2】：

您可以通过多种方式做到这一点。使用 base-r，dplyr 和 data.table 是最典型的。

这是dplyr的方式：

library(dplyr)

df_original %>% group_by(plotID, species) %>% summarize(cover = sum(cover))

#          plotID species     cover
#1 SUF200001035014    ABBA 26.893939
#2 SUF200001035014    BEPA  5.681818
#3 SUF200001035014   PIBA2  9.469697
#4 SUF200001035014    PIMA 16.287879
#5 SUF200001035014    PIRE  1.893939
#6 SUF200046012040   PIBA2 20.454546
#7 SUF200046012040    PIMA 27.651515
#8 SUF200046012040    PIRE 11.363636
#9 SUF200046012040   POTR5 31.439394

这将是 base-r 方式：

aggregate(df_original$cover, by=list(df_original$plotID, df_original$species), FUN=sum)

还有一个data.table的方式——

    library(data.table)
    DT <- as.data.table(df_original)
    DT[, lapply(.SD,sum), by = "plotID,species"]

【参考方案3】：

如上所述，ddply 来自 plyr 包

    library(plyr)
    ddply(df_original, c("plotID","species"), summarise,cover2= sum(cover))


            plotID          species cover2
    1       SUF200001035014 ABBA    26.893939
    2       SUF200001035014 BEPA    5.681818
    3       SUF200001035014 PIBA2   9.469697
    4       SUF200001035014 PIMA    16.287879
    5       SUF200001035014 PIRE    1.893939
    6       SUF200046012040 PIBA2   20.454546
    7       SUF200046012040 PIMA    27.651515
    8       SUF200046012040 PIRE    11.363636
    9       SUF200046012040 POTR5   31.439394