Hello Galaxy-Dev List,

I am having some issues in getting a script to work correctly in Galaxy.  My 
script is written in Python, uses RPy2 to pass some data to the R environment 
and then calls a custom R function; the output of this function, and of the 
greater Python/R tool, is a file.  I am having 2 problems: the first is that R 
cannot find the R file holding the function, and the second is that Galaxy 
cannot find the output file.  Before I elaborate on these issues, I should 
point out that when called from the command line this script executes exactly 
as I intend it to.

First issue:

When called by Galaxy, the Python portion of the script runs fine, and the RPy2 
portion devoted to setting up the R environment so that the script can run 
seems to execute just fine.  The script fails when it tells R to call the 
custom function, with the following output:

Python output:

Traceback (most recent call last):
 line 865, in <module>
    if __name__=='__main__': main()
 line 851, in main
 line 82, in __call__
    return super(SignatureTranslatedFunction, self).__call__(*args, **kwargs)
 line 34, in __call__
    res = super(Function, self).__call__(*new_args, **new_kwargs)
rpy2.rinterface.RRuntimeError: Error in file(file, "r", encoding = encoding) : 
  cannot open the connection

R output:

[1] "/Users/kipperfletez-brant/galaxy-dist/database/job_working_directory/106"

Error in file(file, "r", encoding = encoding) : 
  cannot open the connection
In addition: Warning message:
In file(file, "r", encoding = encoding) :
  cannot open file 
 No such file or directory

The first line of the R output is the return of a call to the Python function 
os.path.abspath('rocprc.R') (note that os.path.realpath('rocprc.R') returns the 
same path), while the second line, for comparison, is a call to R's getwd() - 
the R file is in the same directory as the R environment's setting, but for 
some reason R cannot find the file.  Because this works from the command line, 
I suspect that I am not taking some Galaxy setting into account.

Second issue:

Trying to identify even the kind of error, I hard-coded the directory in which 
I know the R file to reside in a call to R's setwd().  This time, the script 
runs correctly, but there is no file returned by Galaxy.  The script execution 
ends with the following output message:

format: pdf, database: ?
Info: [1] "/Users/kipperfletez-brant/galaxy-dist/tools/myTools/svm_enh"

Watching the file in which the R script is held, I know that the output PDF is 
saved in the same file as that specified in my tool's XML, which, for other 
tools I've written, has resulted in an output file.  Looking at the log, the 
following DEBUG message is written:

galaxy.jobs DEBUG 2012-01-09 10:16:49,899 finish(): Could not move 
 to /Users/kipperfletez-brant/galaxy-dist/database/files/000/dataset_111.dat as 
directed by from_work_dir

As with the first issue, I suspect there is some sort of Galaxy setting I am 
unaware of.

Generally, because both issues relate to finding files not directly 
created/referenced by the caller Python script, I suspect that there will be 
some commonality between the resolutions of the two issues, but I am need help 
to see that.

Thank you for reading,
Kipper Fletez-Brant

Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:


Reply via email to