Question

嗨，我正在尝试将以下XML代码转换为R中的数据帧。但是我不能，因为每条记录缺少值。

RecordID 23063具有以下数据：ActivityCreatedDate，ExpectedInstallDate，InvoiceTxnDate。但是，以下某些节点并不具备所有这些元素。 RecordID 23321缺少InvoiceTxnDate等。

<?xml version="1.0" encoding="windows-1252" ?>
  <Record>
    <RecordID>23063</RecordID>
    <ActivityCreatedDate>2018-12-11T19:00:00</ActivityCreatedDate>
    <ExpectedInstallDate>2018-12-19T19:00:00</ExpectedInstallDate>
    <InvoiceTxnDate>2018-12-13T19:00:00</InvoiceTxnDate>
  </Record>
  <Record>
    <RecordID>23321</RecordID>
    <ActivityCreatedDate>2018-10-15T18:00:00</ActivityCreatedDate>
    <ExpectedInstallDate>2018-11-14T19:00:00</ExpectedInstallDate>
  </Record>
  <Record>
    <RecordID>23566</RecordID>
    <ActivityCreatedDate>2019-01-23T19:00:00</ActivityCreatedDate>
  </Record>
  <Record>
    <RecordID>23217</RecordID>
    <ActivityCreatedDate>2018-12-20T19:00:00</ActivityCreatedDate>
    <ExpectedInstallDate>2019-01-23T19:00:00</ExpectedInstallDate>
    <InvoiceTxnDate>2019-01-18T19:00:00</InvoiceTxnDate>
  </Record>
  <Record>
    <RecordID>23325</RecordID>
    <ActivityCreatedDate>2018-05-25T18:00:00</ActivityCreatedDate>
    <ExpectedInstallDate>2019-01-23T19:00:00</ExpectedInstallDate>
  </Record>
</end of file>

当前我正在使用xml2。我正在使用read_xml将其读取到变量，然后使用xml_find_all和trimws将列存储到列表。然后，我尝试将列表转换为数据框，但由于维度已关闭，因此失败。

我想知道如何将上述XML转换为如下所示的数据框：

RecordID	ActivityCreatedDate	ExpectedInstallDate	InvoiceTxnDate
1	23063	2018-12-11T19:00:00	2018-12-19T19:00:00	2018-12-13T19:00:00
2	23321	2018-10-15T18:00:00	2018-11-14T19:00:00	NA
3	23566	2019-01-23T19:00:00	NA	                NA
4	23217	2018-12-20T19:00:00	2019-01-23T19:00:00	2019-01-18T19:00:00
5	23325	2018-05-25T18:00:00	2019-01-23T19:00:00	NA

在这种情况下，是否有办法遍历每个RecordID并添加

<InvoiceTxnDate>NA</InvoiceTxnDate> or a <ExpectedInstallDate>NA</ExpectedInstallDate>

该节点是否丢失？然后，我会很乐意分享我拥有的统一数据的R代码。另外，如果这个问题没有道理，请告诉我，我将进一步解释自己。

谢谢

Answer 1

您是否尝试过使用XML软件包？

XML::xmlToDataFrame('path to xml file')


> XML::xmlToDataFrame('~/R/test.xml')
  RecordID ActivityCreatedDate ExpectedInstallDate      InvoiceTxnDate
1    23063 2018-12-11T19:00:00 2018-12-19T19:00:00 2018-12-13T19:00:00
2    23321 2018-10-15T18:00:00 2018-11-14T19:00:00                <NA>
3    23566 2019-01-23T19:00:00                <NA>                <NA>
4    23217 2018-12-20T19:00:00 2019-01-23T19:00:00 2019-01-18T19:00:00
5    23325 2018-05-25T18:00:00 2019-01-23T19:00:00                <NA>

如果XML完全如上所示，没有根节点。您可以执行以下操作：

library(xml2)
library(rvest)
library(tidyverse)

## METHOD 1
## add missing root node
read_html('~/R/test.xml') %>% html_children() %>% 
  as_xml_document(root = 'doc') %>% xml_contents() %>% xml_contents() %>% 
  map_df(., function(x) {
    kids <- xml_children(x)
    setNames(as.list(type.convert(xml_text(kids))), xml_name(kids))
  })

## METHOD 2
## treating the xml as a list
read_html('~/R/test.xml') %>% 
  html_nodes('record') %>% 
  as_list() %>% 
  lapply(., function(x) unlist(x, recursive = F) %>% bind_cols()) %>% 
  bind_rows()


## both of the above methods will return the following tibble
# A tibble: 5 x 4
  recordid activitycreateddate expectedinstalldate invoicetxndate     
  <chr>    <chr>               <chr>               <chr>              
1 23063    2018-12-11T19:00:00 2018-12-19T19:00:00 2018-12-13T19:00:00
2 23321    2018-10-15T18:00:00 2018-11-14T19:00:00 NA                 
3 23566    2019-01-23T19:00:00 NA                  NA                 
4 23217    2018-12-20T19:00:00 2019-01-23T19:00:00 2019-01-18T19:00:00
5 23325    2018-05-25T18:00:00 2019-01-23T19:00:00 NA

在R中将XML转换为数据帧

1 个答案: