通过 R 中的组 id 将 dataframe 转换为矩阵格式

Question

I'm trying to convert a data.frame object into matrix format by some group id using the dlply function, but the resulting object clearly is not of matrix format. Could anyone point out what's going on here?

 library(plyr) #load the data export <- readRDS(url("https://www.dropbox.com/s/7nn4kcdbsiteplj/export.rds?dl=1")) export_mat <- dlply(export,.(reporter_iso), fun = as.matrix) str(export_mat) List of 161 $ afg:'data.frame': 161 obs. of 6 variables: ..$ reporter_iso: chr [1:161] "afg" "afg" "afg" "afg"... ..$ partner_iso: chr [1:161] "afg" "ago" "alb" "are"... ..$ cong_pct: num [1:161] 0 0 0 0.0915 0... ..$ cag_pct: num [1:161] 0 0 0 0.0297 0... ..$ ig_pct: num [1:161] 0 0 0 0.049 0... ..$ rm_pct: num [1:161] 0 0 0 0.83 0... ..- attr(*, "vars")= chr "reporter_iso" $ ago:'data.frame': 161 obs. of 6 variables: ..$ reporter_iso: chr [1:161] "ago" "ago" "ago" "ago"... ..$ partner_iso: chr [1:161] "afg" "ago" "alb" "are"... ..$ cong_pct: num [1:161] 0 0 0 0.203 0... ..$ cag_pct: num [1:161] 0 0 0 0.698 0... ..$ ig_pct: num [1:161] 0 0 0 0.0148 0... ..$ rm_pct: num [1:161] 0 0 0 0.0842 1... ..- attr(*, "vars")= chr "reporter_iso"

Answer 1

Wrong argument name, try this:

export_mat <- dlply(export,.(reporter_iso), .fun = as.matrix)

Answer 2

In base R, you can use by :

export_mat <- by(export, export$reporter_iso, as.matrix)

Or split + lapply :

export_mat <- lapply(split(export, export$reporter_iso), as.matrix)

A matrix can hold data of only one type in it. Since you have characters as well as numbers in your data all the data is transformed to character.