MPO 581 Class number 10/27 Feb 21, 2011
Topic: Histograms (1D and 2D) and composites (modes? or just tails?)
Loose ends from last class:
Teams
and
homework
leaders
all
assigned!?
(edit amongst
yourselves)
Testable questions
compilation taking shape: add to it, for course brownie points.
Today's material:
Histograms and 2d histograms:
what's their point?
(showed powerpoint of a few research
things mostly mine that illustrate concepts)
- If you find a robust, repeatable bimodal (or multimodal)
distribution, that's gold: a qualitative
distinction in a world of continuum mush.
- Even if you find a mushy unimodal thing, seeing it is guidance
for where to slice the tails.
Choosing modes (ideally), or
quantiles (slices of the tails
of a unimodal thing) as a basis for
composites
- You can do it manually: with find() in Matlab or where() in IDL
or Python.
- Fancy histogram programs may give you back an array of indices
pointing to which values fall in each bin.
Composites (conditional averages)
- Line up all the values meeting your criteria and average them
- You can average fields other than the one that defined the
criteria.
- Example: composite weather maps around tropical cyclone events.
Can be
HW2 - Multivariate Mayhem!
datasets and tools updated.
Open questions, assignments, and
loose
ends for next class:
Testable questions about today's
material: