aflaxman

import pymc as pm
import numpy as np

# FIXME: Need to store duplicates too, when jumps are rejected. That means some mechanism
# for making sure the history is full-rank needs to be employed.

class HistoryCovarianceStepper(pm.StepMethod):
    
    _state = ['n_points','history','tally','verbose']
    

aflaxman

aflaxman

*~

aflaxman

import IPython.core.display
import matplotlib as mpl

def clean():
    # set mpl defaults for nice display
    mpl.rcParams['font.size'] = 12
    mpl.rcParams['figure.figsize'] = (18, 6)
    mpl.rcParams['lines.linewidth'] = 1

aflaxman

{
 "metadata": {
  "name": "2013_04_20b_hello_d3-vega-vincent"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {

aflaxman

Interactive graphic by Christopher Ingraham.

Source: American Community Survey 2012.

Last week this graphic of migration flows by Chris Walker made the rounds. Many commenters noted how beautiful the graphic is, and rightly so.

I always have a bit of a hard time parsing exactly what's going on within these Circos-style visualizations. Walker's graphic allows you to hover over a given state to highlight only those migrations, which helps quite a bit.

I thought there might be a better way to display these data, but I wasn't right. I stuck with a map, drawing circles for each state sized by net migration (comings minus goings) and colored according to whether the state gained or lost residents overall. To get at individual state flows, click a state - paths radiate inwards and outwards fr

aflaxman

{
 "metadata": {
  "name": "",
  "signature": "sha256:53400d77ba7c32c0cc425b9c26b94a2bf7c3ed3c199a7b07c9c6526e9d844d8d"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [