TypeError for built-in operation

edit

In working with tf.Transformer, I've run into this error many times.

TypeError: bad argument type for built-in operation

The most frustrating part of getting this message is that the traceback blames seemingly random lines of code for the error. Here's one reduced way to get the error.

#!/usr/bin/python

import roslib; roslib.load_manifest('tf')
import tf, rospy
from geometry_msgs.msg import TransformStamped

if __name__ == '__main__':
    tfman = tf.Transformer(True, rospy.Duration(10.0))
    angle = tf.transformations.quaternion_from_euler(0,0,0)

    m = TransformStamped()
    m.header.frame_id = 'root'
    m.child_frame_id = 'child'
    print type(angle), type(m.transform.rotation)
    m.transform.rotation = angle
    tfman.setTransform(m)

    for i in range(1):
        print i

This outputs:

<type 'numpy.ndarray'=""> <class 'geometry_msgs.msg._quaternion.quaternion'="">
0
Traceback (most recent call last):
  File "src/error.py", line 18, in <module>
    for i in range(1):
TypeError: bad argument type for built-in operation

The clearest way to fix this is to set the rotation using a quaternion, not a numpy array. However, is there some way to get a clearer error message?

Edit: I guess the clearer question is: Why would this mistake cause a TypeError on line 18? Also, I put the error here so that the next person who searches for the error message on the board will find it here.