R imports XML from Graphpad Prism
I am trying to import raw data from Graphpad Prism.pzfx files, which are basically .xml files. I took out most of the specific items and left only the part that interests me.
<?xml version="1.0" encoding="UTF-8"?>
<GraphPadPrismFile xmlns="http://graphpad.com/prism/Prism.htm" PrismXMLVersion="5.00">
<TableSequence Selected="1">
<Ref ID="Table0" Selected="1"/>
</TableSequence>
<Table ID="Table0" XFormat="error" YFormat="replicates" Replicates="1" TableType="XY" EVFormat="AsteriskAfterNumber">
<Title>Data 1</Title>
<XColumn Width="162" Decimals="0" Subcolumns="1">
<Title>X</Title>
<Subcolumn>
<d>1</d>
<d>2</d>
<d>3</d>
<d>4</d>
<d>5</d>
</Subcolumn>
</XColumn>
<YColumn Width="81" Decimals="4" Subcolumns="1">
<Title>ML</Title>
<Subcolumn>
<d>120</d>
<d>100</d>
<d>5</d>
<d>0</d>
<d>1.5</d>
</Subcolumn>
</YColumn>
<YColumn Width="81" Decimals="4" Subcolumns="1">
<Title>MH</Title>
<Subcolumn>
<d>10</d>
<d>560</d>
<d>665</d>
<d>40</d>
<d>31.5</d>
</Subcolumn>
</YColumn>
<YColumn Width="81" Decimals="6" Subcolumns="1">
<Title>MH2</Title>
<Subcolumn>
<d>1.20</d>
<d>100</d>
<d>5</d>
<d>0</d>
<d>1.5</d>
</Subcolumn>
</YColumn>
<YColumn Width="81" Decimals="6" Subcolumns="1">
<Title>MH1</Title>
<Subcolumn>
<d>120</d>
<d>100</d>
<d>5</d>
<d>0</d>
<d>1.5</d>
</Subcolumn>
</YColumn>
</Table>
</GraphPadPrismFile>
As I understand it, I have a Node table that again has XColumn, YColumn nodes that have a Title and then have subcolumn nodes. They contain my raw data, which I would like to convert to R-data.frame.
So far I have managed to import one YColumn: (file is the path to the above)
xData <- xmlParse(file)
xData.rt <- xmlRoot(xData)
xmlToDataFrame(xData.rt[["Table"]][["YColumn"]][["Subcolumn"]])
This gives me the first YColumn as data.frame:
text
1 120
2 100
3 5
4 0
5 1.5
Is there a way (maybe easier?) To import all X and Y Colums into one data.frame?
source to share
Here are some ideas ...
This finds all the nodes XColumn
and YColumn
:
columns = xpathSApply(xData,"//ns:XColumn|//ns:YColumn",
namespaces=c(ns="http://graphpad.com/prism/Prism.htm"))
Then, given one of them, it gets the title and values:
> xpathSApply(columns[[2]],".//ns:Title",
namespaces=c(ns="http://graphpad.com/prism/Prism.htm"),xmlValue)
[1] "ML"
> xpathSApply(columns[[2]],".//ns:d",
namespaces=c(ns="http://graphpad.com/prism/Prism.htm"),xmlValue)
[1] "120" "100" "5" "0" "1.5"
You can find out if there is something XColumn
or YColumn
by checking it xmlName
:
> xmlName(columns[[1]])
[1] "XColumn"
> xmlName(columns[[2]])
[1] "YColumn"
This is probably enough building kit pieces. Loop over columns
and sum the values using cbind
probably ...
Obviously character based numbers need conversion with as.numeric
, and I'm not entirely sure how you want to combine the values XColumn
and YColumn
given there a smaller value XColumn
...
source to share