R：如何在“ by”是变量的函数中使用聚合？

Question

I'm trying to write a function that creates new data frames with summarised columns. I can use aggregate individually:

df = aggregate(. ~ Family, data=df, FUN=sum)

But I want to replace the 'Family' with an object from a list:

levels = c('Family','Order','Class','Phylum','Superkingdom')
for (i in 1:5){
    OTU = levels[i]
    df = lineage[,c(OTU,'Abundance')]
    df = aggregate(. ~ OTU, data=df, FUN=sum)
}

In the code above, 'lineage' in a data frame with the taxonomic levels for a selection of organisms, plus an abundance value. What I'd like to do is create a new, summarised, data frame for each taxonomic level, with the abundance values being summarised.

I've found many posts about aggregate on here, but still can't get it to work.

Thanks in advance for any help!

Answer 1

I include two variables to summarise in case you're interested in extending it beyond simply Abundance . Something like this could work:

set.seed(42)

taxa <- c('Family','Order','Class','Phylum','Superkingdom')
df <- data.frame(ID = 1:15, Abundance = runif(15), HomeRange = abs(rnorm(15)), taxa = rep(taxa, times = 3))

(output2 <- aggregate(cbind(HomeRange, Abundance) ~ taxa, data = df, FUN = sum))

With a result like this:

#           taxa HomeRange Abundance
# 1        Class  1.413938  1.355478
# 2       Family  2.556280  1.891644
# 3        Order  3.317171  2.392776
# 4       Phylum  2.129622  1.742869
# 5 Superkingdom  4.031149  1.809103

Similarly, for one variable:

(output1 <- aggregate(Abundance ~ taxa, data = df, FUN = sum))

#           taxa Abundance
# 1        Class  1.355478
# 2       Family  1.891644
# 3        Order  2.392776
# 4       Phylum  1.742869
# 5 Superkingdom  1.809103