xml parser in R with hierarchical nodes, tags and values

问题

I am trying to parse sample_attributes (preferably all) from the following xml file. Tried a couple of things but the XML gets clumped into one node:

xml.url <- "http://www.ebi.ac.uk/ena/data/view/ERS445758&display=xml"
xmlfile <- xmlTreeParse(xml.url)
xmltop = xmlRoot(xmlfile)
IBDcat <- xmlSApply(xmltop, function(x) xmlSApply(x, xmlValue))

Also tried solutions mentioned here: How to parse XML to R data frame and how to create an R data frame from a xml file but when I try something like:

data <- xmlParse("http://www.ebi.ac.uk/ena/data/view/ERS445758&display=xml")
xml_data <- xmlToList(data)
xmlToDataFrame(nodes=getNodeSet(data,"/SAMPLE_ATTRIBUTE"))[c("age","sex","body site","body-mass index")]

I get an error saying undefined columns selected

Any help will be appreciated thanks!

回答1:

At least for your second attempt, you just needed to select any SAMPLE_ATTRIBUTE node using //. Then subset by tag.

doc <- xmlParse(xml.url)
x <- xmlToDataFrame(getNodeSet(doc,"//SAMPLE_ATTRIBUTE"))
## OR 
xmlToDataFrame(doc["//SAMPLE_ATTRIBUTE"])
                  TAG      VALUE UNITS
1  investigation type metagenome  <NA>
2        project name       BMRP  <NA>
3 experimental factor microbiome  <NA>
4         target gene   16S rRNA  <NA>
5  target subfragment       V1V2  <NA>
...


subset(x, TAG %in% c("age","sex","body site","body-mass index") )
               TAG         VALUE UNITS
15             age            28 years
16             sex          male  <NA>
17       body site Sigmoid colon  <NA>
19 body-mass index    16.9550173  <NA>

回答2:

Here's a tidyverse option; xml2 has a simple read_xml function that has an associated as_list function. purrr is a package for manipulating lists that is very handy, though you could, of course, do the same things in base R if you prefer.

library(xml2)
library(purrr)

x <- read_xml("http://www.ebi.ac.uk/ena/data/view/ERS445758&display=xml")

x_list <- as_list(x)

x_df <- x_list %>% map('SAMPLE_ATTRIBUTES') %>% flatten() %>% map_df(flatten)

x_df
#> # A tibble: 35 × 3
#>                       TAG                               VALUE UNITS
#>                     <chr>                               <chr> <chr>
#> 1      investigation type                          metagenome  <NA>
#> 2            project name                                BMRP  <NA>
#> 3     experimental factor                          microbiome  <NA>
#> 4             target gene                            16S rRNA  <NA>
#> 5      target subfragment                                V1V2  <NA>
#> 6             pcr primers                            27F-338R  <NA>
#> 7   multiplex identifiers                          TGATACGTCT  <NA>
#> 8       sequencing method                      pyrosequencing  <NA>
#> 9  sequence quality check                            software  <NA>
#> 10          chimera check ChimeraSlayer; Usearch 4.1 database  <NA>
#> # ... with 25 more rows

or do the subsetting in XPath instead:

x %>% xml_find_all('//SAMPLE_ATTRIBUTE') %>% map(as_list) %>% map_df(flatten)

which returns the same thing.

回答3:

Slightly different approach to the very creative one by @allistaire:

library(xml2)

doc <- read_xml("http://www.ebi.ac.uk/ena/data/view/ERS445758&display=xml")

xml_find_all(doc, ".//SAMPLE_ATTRIBUTE") %>% 
  map(xml_children) %>% 
  map_df(~as.list(setNames(xml_text(.), xml_name(.))))

来源：https://stackoverflow.com/questions/41007689/xml-parser-in-r-with-hierarchical-nodes-tags-and-values

标签

xml

xml-parsing