[R] XML: Reading transition matrices into R

[Duncan Temple Lang]

stefan.duke at gmail.com wrote:
> Hello,
> thanks a lot. This is a form which I can work with in R.
> 
> Another question, which I hope is not a bridge too far: Can I write
> the R matrix which your code created back into the xml format (i.e.
> with the same tags and structure) from which it came and hence feed it
> back to the original software?


trans = apply(xx, 1, function(x) {
                       tr = newXMLNode("transition")
                       mapply(newXMLNode, names(x), x, MoreArgs = list(parent = tr))
                       tr
                     })
top = newXMLNode("transitionmatrix", .children = trans)
saveXML(top, "newTransition.xml")




> 
> Best,
> Stefan
> 
> 
> On Thu, Nov 12, 2009 at 3:17 PM, Duncan Temple Lang
> <duncan at wald.ucdavis.edu> wrote:
>>
>> stefan.duke at gmail.com wrote:
>>> Hello,
>>> from a software I have the following output in xml (see below):
>>> It is a series of matrices, for each age one. I have 3 categories
>>> (might vary in the application), hence, 3x3 matrices where each
>>> element gives the probability of transition from i to j. I would like
>>> read this data into R (preferably in a list, where each list element
>>> is one of the age specific matrices) and - after altering the values
>>> in R - write it back into the file.  I know that there is an xml
>>> package in R with which I have already struggled, but I have to admit
>>> my understanding is too limited. Maybe somebody had a similar problem
>>> or know the code of the top of his or her head.
>> Hi Stefan
>>
>>  There are many approaches for handling this. I assume that the primary
>> obstacle you are facing is extracting the values from the XML.  The following
>> will do that for you.
>> We start with the content in transition.xml (or in a string in R).
>> Since the XML is very shallow, i.e. not very hierarchical, and all
>> the information is in the transition nodes under the root, we can
>> use xmlToList().
>> This returns a list with an element for each <transition>
>> element, and such elements are character vectors containing the
>> values from age, sex, from, to, and percent.
>> So I've put these into a matrix and you are now back entirely
>> within R and can group the values by age and arrange them into
>> the individual transition matrices.
>>
>>
>>   doc = xmlParse("transition.xml")
>>   matrix(as.numeric(unlist(xmlToList(doc))), , 5, byrow = TRUE,
>>          dimnames = list(NULL, names(xmlRoot(doc)[[1]])))
>>
>>
>>  D.
>>
>>
>>
>>> Any help appreciated.
>>>
>>> Thanks and best,
>>> Stefan
>>>
>>>
>>>
>>>
>>>
>>> <?xml version="1.0" encoding="UTF-8" standalone="no"?><transitionmatrix>
>>> <transition><age>0</age><sex>0</sex><from>1</from><to>1</to><percent>99.99999</percent></transition><transition><age>0</age><sex>0</sex><from>1</from><to>2</to><percent>0.0</percent></transition><transition><age>0</age><sex>0</sex><from>1</from><to>3</to><percent>0.0</percent></transition><transition><age>0</age><sex>0</sex><from>2</from><to>1</to><percent>0.0</percent></transition><transition><age>0</age><sex>0</sex><from>2</from><to>2</to><percent>99.999985</percent></transition><transition><age>0</age><sex>0</sex><from>2</from><to>3</to><percent>0.0</percent></transition><transition><age>0</age><sex>0</sex><from>3</from><to>1</to><percent>0.0</percent></transition><transition><age>0</age><sex>0</sex><from>3</from><to>2</to><percent>0.0</percent></transition><transition><age>0</age><sex>0</sex><from>3</from><to>3</to><percent>99.999985</percent></transition><transition><age>0</age><sex>1</sex><from>1</from><to>1</to><percent>100.0</percent></transition><transition><age>
0<
>> /!
>>>  age><sex>1</sex><from>1</from><to>2</to><percent>0.0</percent></transition><transition><age>0</age><sex>1</sex><from>1</from><to>3</to><percent>0.0</percent></transition><transition><age>0</age><sex>1</sex><from>2</from><to>1</to><percent>0.0</percent></transition><transition><age>0</age><sex>1</sex><from>2</from><to>2</to><percent>100.0</percent></transition><transition><age>0</age><sex>1</sex><from>2</from><to>3</to><percent>0.0</percent></transition><transition><age>0</age><sex>1</sex><from>3</from><to>1</to><percent>0.0</percent></transition><transition><age>0</age><sex>1</sex><from>3</from><to>2</to><percent>0.0</percent></transition><transition><age>0</age><sex>1</sex><from>3</from><to>3</to><percent>100.0</percent></transition><transition><age>1</age><sex>0</sex><from>1</from><to>1</to><percent>99.99999</percent></transition><transition><age>1</age><sex>0</sex><from>1</from><to>2</to><percent>0.0</percent></transition><transition><age>1</age><sex>0</sex><from>1</f
ro
>> m!
>>>  ><to>3</to><percent>0.0</percent></transition><transition><age>1</age>
>>> <sex>0</sex><from>2</from><to>1</to><percent>0.0</percent></transition><transition><age>1</age><sex>0</sex><from>2</from><to>2</to><percent>99.999985</percent></transition><transition><age>1</age><sex>0</sex><from>2</from><to>3</to><percent>0.0</percent></transition><transition><age>1</age><sex>0</sex><from>3</from><to>1</to><percent>0.0</percent></transition><transition><age>1</age><sex>0</sex><from>3</from><to>2</to><percent>0.0</percent></transition><transition><age>1</age><sex>0</sex><from>3</from><to>3</to><percent>99.999985</percent></transition><transition><age>1</age><sex>1</sex><from>1</from><to>1</to><percent>100.0</percent></transition><transition><age>1</age><sex>1</sex><from>1</from><to>2</to><percent>0.0</percent></transition><transition><age>1</age><sex>1</sex><from>1</from><to>3</to><percent>0.0</percent></transition><transition><age>1</age><sex>1</sex><from>2</from><to>1</to><percent>0.0</percent></transition><transition><age>1</age><sex>1</sex><from>2</f
ro
>> m!
>>>  ><to>2</to><percent>100.0</percent></transition><transition><age>1</age><sex>1</sex><from>2</from><to>3</to><percent>0.0</percent></transition><transition><age>1</age><sex>1</sex><from>3</from><to>1</to><percent>0.0</percent></transition><transition><age>1</age><sex>1</sex><from>3</from><to>2</to><percent>0.0</percent></transition><transition><age>1</age><sex>1</sex><from>3</from><to>3</to><percent>100.0</percent></transition><transition><age>2</age><sex>0</sex><from>1</from><to>1</to><percent>99.35205</percent></transition><transition><age>2</age><sex>0</sex><from>1</from><to>2</to><percent>0.6479474</percent></transition><transition><age>2</age><sex>0</sex><from>1</from><to>3</to><percent>0.0</percent></transition><transition><age>2</age><sex>0</sex><from>2</from><to>1</to><percent>0.0</percent></transition><transition><age>2</age><sex>0</sex><from>2</from><to>2</to><percent>97.101456</percent></transition><transition><age>2</age><sex>0</sex><from>2</from><to>3</to><pe
rc
>> e!
>>>  nt>2.8985496</percent></transition><transition><age>2</age><sex>0</sex
>>>> <from>3</from><to>1</to><percent>0.0</percent></transition><transition><age>2</age><sex>0</sex><from>3</from><to>2</to><percent>0.0</percent></transition><transition><age>2</age><sex>0</sex><from>3</from><to>3</to><percent>100.0</percent></transition><transition><age>2</age><sex>1</sex><from>1</from><to>1</to><percent>97.82136</percent></transition><transition><age>2</age><sex>1</sex><from>1</from><to>2</to><percent>2.1786473</percent></transition><transition><age>2</age><sex>1</sex><from>1</from><to>3</to><percent>0.0</percent></transition><transition><age>2</age><sex>1</sex><from>2</from><to>1</to><percent>0.0</percent></transition><transition><age>2</age><sex>1</sex><from>2</from><to>2</to><percent>93.33336</percent></transition><transition><age>2</age><sex>1</sex><from>2</from><to>3</to><percent>6.6666408</percent></transition><transition><age>2</age><sex>1</sex><from>3</from><to>1</to><percent>0.0</percent></transition><transition><age>2</age><sex>1</sex><from>3</fr
om
>>> !
>>>  <to>2</to><percent>0.0</percent></transition><transition><age>2</age><sex>1</sex><from>3</from><to>3</to><percent>100.0</percent></transition>
>>> </transitionmatrix>
>>>
>>> ______________________________________________
>>> R-help at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.