[jira] [Created] (SPARK-17587) SparseVector __getitem__ should follow __getitem__ contract

["Maciej Szymkiewicz (JIRA)" <ji...@apache.org>]

Maciej Szymkiewicz created SPARK-17587:
------------------------------------------

             Summary: SparseVector __getitem__ should follow __getitem__ contract
                 Key: SPARK-17587
                 URL: https://issues.apache.org/jira/browse/SPARK-17587
             Project: Spark
          Issue Type: Bug
          Components: ML, MLlib, PySpark
    Affects Versions: 2.0.0, 1.6.0
            Reporter: Maciej Szymkiewicz
            Priority: Minor


According to {{\_\_getitem\_\_}} [contract|https://docs.python.org/3/reference/datamodel.html#object.__getitem__]:

{quote}
if of a value outside the set of indexes for the sequence (after any special interpretation of negative values), {{IndexError}} should be raised.
{quote}

This required for example for correct iteration over the structure.

Right now it throws {{ValueError}} what results in a quite confusing behavior when attempt to iterate over a vector results in a {{ValueError}} due to unterminated iteration:

{code}

In [1]: from pyspark.mllib.linalg import SparseVector

In [2]: list(SparseVector(4, [0], [0]))
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-2-147f3bb0a47d> in <module>()
----> 1 list(SparseVector(4, [0], [0]))

/opt/spark-2.0/python/pyspark/mllib/linalg/__init__.py in __getitem__(self, index)
    803 
    804         if index >= self.size or index < -self.size:
--> 805             raise ValueError("Index %d out of bounds." % index)
    806         if index < 0:
    807             index += self.size

ValueError: Index 4 out of bounds.
{code}





--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org