What is the best way to make nested list with large datasets?? R-CodePudding

What is the best way to make nested lists with large datasets. Currently I am adding the elements with a for but I don't know if it is the most efficient way to do it.

In the example below I want to fill a nested list with the taxonomic levels Domain> Kingdom> Phylum> Class> Order> Family> Genus> Species.

So the code currently goes through each taxonomic level collecting the information and filling in the list, but due to the amount of data, the process takes too long and I would like to know if there is a way to optimize the process.

The code is attached below. I appreciate any suggestions, comments, etc.

Thank you


data <- data.frame(Reino   = c("reinoa","reinoa","reinob","reinoc"),
                   Filo    = c("Filoa1","Filoa2","Filob","Filoc"),
                   Clase   = c("Clasea1","Clase2","Claseb","Clasec"),
                   Orden   = c("Ordena1","Ordena2","Ordenb","Ordenc"),
                   Familia = c("Familiaa1","Familiaa2","Familiab","Familiac"),
                   Genero  = c("Generoa1","Generoa2","Generob","Generoc"),
                   Especie = c("Especiea1","Especiea2","Especieb","Especiec"))

for(i in unique(data$Reino)){
  
  dftaxonomica[[i]] <- list()
  
  print(i)
  
  for(j in unique(data[data$Reino==i,]$Filo)){
    
    dftaxonomica[[i]][[j]] <- list()
    
    for(w in unique(data[data$Reino==i & data$Filo==j,]$Clase)){
      
      dftaxonomica[[i]][[j]][[w]] <- list()
      
      
      for(z in unique(data[data$Reino==i & data$Filo==j & data$Clase == w,]$Orden)){
        
        dftaxonomica[[i]][[j]][[w]][[z]] <- list()
        
        for(h in unique(data[data$Reino==i & data$Filo==j & data$Clase == w & data$Orden == z,]$Familia)){
          
          dftaxonomica[[i]][[j]][[w]][[z]][[h]] <- list()
          
          for(q in unique(data[data$Reino==i & data$Filo==j & data$Clase == w & data$Orden == z & data$Familia == h,]$Genero)){
            
            dftaxonomica[[i]][[j]][[w]][[z]][[h]][[q]] <- list()
            
            for(k in unique(data[data$Reino==i & data$Filo==j & data$Clase == w & data$Orden == z & data$Familia == h & data$Género == h,]$Especie)){
              
              dftaxonomica[[i]][[j]][[w]][[z]][[h]][[q]][[k]] <- list()
              
              
              
              
            }
            
            
          } 
          
          
        }  
        
        
        
        
      }  
      
      
    }
    
    
    
    
    
  }
  
}

CodePudding user response：

A recursive function might help:

recurse <- function(x) {
    nms <- names(x)
    if (length(nms) > 1L) {
        lapply(split(x[nms[-1L]], x[[nms[1L]]]), recurse)
    } else {
        nms <- unique(x[[1L]])
        setNames(vector("list", length(nms)), nms)
    }
}

recurse(data)

$reinoa
$reinoa$Filoa1
$reinoa$Filoa1$Clasea1
$reinoa$Filoa1$Clasea1$Ordena1
$reinoa$Filoa1$Clasea1$Ordena1$Familiaa1
$reinoa$Filoa1$Clasea1$Ordena1$Familiaa1$Generoa1
$reinoa$Filoa1$Clasea1$Ordena1$Familiaa1$Generoa1$Especiea1
NULL

$reinoa$Filoa2
$reinoa$Filoa2$Clase2
$reinoa$Filoa2$Clase2$Ordena2
$reinoa$Filoa2$Clase2$Ordena2$Familiaa2
$reinoa$Filoa2$Clase2$Ordena2$Familiaa2$Generoa2
$reinoa$Filoa2$Clase2$Ordena2$Familiaa2$Generoa2$Especiea2
NULL


$reinob
$reinob$Filob
$reinob$Filob$Claseb
$reinob$Filob$Claseb$Ordenb
$reinob$Filob$Claseb$Ordenb$Familiab
$reinob$Filob$Claseb$Ordenb$Familiab$Generob
$reinob$Filob$Claseb$Ordenb$Familiab$Generob$Especieb
NULL


$reinoc
$reinoc$Filoc
$reinoc$Filoc$Clasec
$reinoc$Filoc$Clasec$Ordenc
$reinoc$Filoc$Clasec$Ordenc$Familiac
$reinoc$Filoc$Clasec$Ordenc$Familiac$Generoc
$reinoc$Filoc$Clasec$Ordenc$Familiac$Generoc$Especiec
NULL

CodePudding user response：

The rrapply() function in the rrapply-package has an (efficient) option how = "unmelt" exactly for this purpose:

library(rrapply)

## add value column
data[["Value"]] <- replicate(nrow(data), list())

## unmelt data.frame
out <- rrapply(data, how = "unmelt")

str(out)
#> List of 3
#>  $ reinoa:List of 2
#>   ..$ Filoa1:List of 1
#>   .. ..$ Clasea1:List of 1
#>   .. .. ..$ Ordena1:List of 1
#>   .. .. .. ..$ Familiaa1:List of 1
#>   .. .. .. .. ..$ Generoa1:List of 1
#>   .. .. .. .. .. ..$ Especiea1: list()
#>   ..$ Filoa2:List of 1
#>   .. ..$ Clase2:List of 1
#>   .. .. ..$ Ordena2:List of 1
#>   .. .. .. ..$ Familiaa2:List of 1
#>   .. .. .. .. ..$ Generoa2:List of 1
#>   .. .. .. .. .. ..$ Especiea2: list()
#>  $ reinob:List of 1
#>   ..$ Filob:List of 1
#>   .. ..$ Claseb:List of 1
#>   .. .. ..$ Ordenb:List of 1
#>   .. .. .. ..$ Familiab:List of 1
#>   .. .. .. .. ..$ Generob:List of 1
#>   .. .. .. .. .. ..$ Especieb: list()
#>  $ reinoc:List of 1
#>   ..$ Filoc:List of 1
#>   .. ..$ Clasec:List of 1
#>   .. .. ..$ Ordenc:List of 1
#>   .. .. .. ..$ Familiac:List of 1
#>   .. .. .. .. ..$ Generoc:List of 1
#>   .. .. .. .. .. ..$ Especiec: list()

Reference link

https://jorischau.github.io/rrapply/articles/articles/2-efficient-melting-unnesting.html#efficient-unmelting-of-melted-data-frames