tf transformpose with Odometry message

edit

I have some Odometry messages arriving in a frame, that I'd like to translate to another frame. I'm trying to do this with

t = tf.TransformerROS(True, rospy.Duration(10.0))
pose_base = t.transformPose("map", msg.pose)

However I get #AttributeError: 'PoseWithCovariance' object has no attribute 'header'

This I think is because the Odometry message itself is stamped with msg.header.stamp, but PoseWithCovarience part, msg.pose, does not have it's own stamp. What's the best way to send this information to tf to get the transformed pose? (I don't care about the twist, but it would be nice to have the covarience of the pose transformed.)