You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@mahout.apache.org by Grant Ingersoll <gs...@apache.org> on 2009/08/17 17:50:05 UTC

Taste FileDataModel changes

I have a file containing:  userId, itemId, float

Prior to the "big" changes in Taste, this loaded just fine.  Now, I'm  
getting:
  [java] Exception in thread "main" java.lang.NumberFormatException:  
For input string: "3135415,4.82"
      [java]     at  
java 
.lang.NumberFormatException.forInputString(NumberFormatException.java: 
48)
      [java]     at java.lang.Long.parseLong(Long.java:412)
      [java]     at java.lang.Long.parseLong(Long.java:461)
      [java]     at  
org 
.apache 
.mahout 
.cf 
.taste 
.impl.model.file.FileDataModel.processLineWithoutID(FileDataModel.java: 
294)
      [java]     at  
org 
.apache 
.mahout 
.cf 
.taste 
.impl.model.file.FileDataModel.processFileWithoutID(FileDataModel.java: 
272)
      [java]     at  
org 
.apache 
.mahout 
.cf.taste.impl.model.file.FileDataModel.buildModel(FileDataModel.java: 
153)
      [java]     at  
org 
.apache 
.mahout 
.cf.taste.impl.model.file.FileDataModel.reload(FileDataModel.java:128)
      [java]     at  
org 
.apache 
.mahout 
.cf.taste.impl.model.file.FileDataModel.checkLoaded(FileDataModel.java: 
311)
      [java]     at  
org 
.apache 
.mahout 
.cf.taste.impl.model.file.FileDataModel.getNumUsers(FileDataModel.java: 
357)
...
      [java] Java Result: 1

It looks like it thinks the file doesn't have IDs (in buildModel)  
because of line 142 in FileDataModel:
boolean hasPrefValues = firstLine.indexOf(delimiter,  
firstLine.indexOf(delimiter) + 1) >= 0;

But, in my file, the first line is a comment.  Here's what my file  
looks like:
#comments go here
#more comments here
990,3135415,4.82
991,3135415,4.79
992,3135415,4.99
993,3135415,4.99
994,3135415,4.72

Presumably, the fix here is to iterate to the first non-comment line,  
right?

-Grant

Re: Taste FileDataModel changes

Posted by Sean Owen <sr...@gmail.com>.

Good call, I'll fix that shortly.

On Mon, Aug 17, 2009 at 4:50 PM, Grant Ingersoll<gs...@apache.org> wrote:
> It looks like it thinks the file doesn't have IDs (in buildModel) because of
> line 142 in FileDataModel:
> boolean hasPrefValues = firstLine.indexOf(delimiter,
> firstLine.indexOf(delimiter) + 1) >= 0;