vector needs clamping scaletype

[danlipsa]

Besides the list of options already available for scaletype, we need to add clamping.
So instead of mapping [minnorm, maxnorm] to a range of colors we map a user specified
[scalerange[0], scalerange[1]].

This will give us the ability to ignore possible outliers in the vector data.

[danlipsa]

@scottwittenburg @doutriaux1 @aashish24 Please comment.

[danlipsa]

See the comment in vector.py scaletype for the available scaling options.

[doutriaux1]

@danlipsa sounds good, what is it doing currently?

[danlipsa]

This is the help for scaletype. The default is constantNNormalize.

      off - No scaling is performed on the vector values
      constant: vector value *  self.scale
      normalize: vector value /  max_norm
      constantNNormalize: vector value * self.scale / max_norm
      linear: map [min_norm, max_norm] to self.scalerange
      constantNLinear - map [min_norm, max_norm] to self.scalerange and then multiply by self.scale