AustinRochford

AustinRochford

" Pathogen
filetype off " Pathogen needs to run before plugin indent on
call pathogen#infect('bundle/{}')
call pathogen#helptags() " generate helptags for everything in 'runtimepath'
filetype plugin indent on

" Colors
colorscheme manxome
syntax on
hi StatusLine ctermfg=white ctermbg=blue

AustinRochford

{
 "metadata": {
  "name": "",
  "signature": "sha256:e9943648b5982fa868ab077d78fe550231c16f916d699ee244cf22a1a6120489"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [

AustinRochford

AustinRochford

{"nbformat_minor": 0, "cells": [{"execution_count": 1, "cell_type": "code", "source": "%matplotlib inline", "outputs": [], "metadata": {"collapsed": true, "trusted": true}}, {"source": "---\ntitle: Maximum Likelihood Estimation of Custom Models in Python with StatsModels\ntags: Statistics, Python\n---", "cell_type": "markdown", "metadata": {}}, {"source": "Maximum likelihood estimation is a common method for fitting statistical models.  In Python, it is quite possible to fit maximum likelihood models using just [`scipy.optimize`](http://docs.scipy.org/doc/scipy/reference/optimize.html). Over time, however, I have come to prefer the convenience provided by [`statsmodels`'](http://statsmodels.sourceforge.net/) [`GenericLikelihoodModel`](http://statsmodels.sourceforge.net/devel/dev/generated/statsmodels.base.model.GenericLikelihoodModel.html).  In this post, I will show how easy it is to subclass `GenericLikelihoodModel` and take advantage of much of `statsmodels`' well-developed machinery for maximum likelihood

AustinRochford

{"nbformat_minor": 0, "cells": [{"execution_count": 1, "cell_type": "code", "source": "%matplotlib inline", "outputs": [], "metadata": {"collapsed": true, "trusted": false}}, {"source": "---\ntitle: Robust Regression with t-Distributed Residuals\ntags: Statistics, PyMC\n---", "cell_type": "markdown", "metadata": {}}, {"source": "Ordinarly least squares (OLS) is, without a doubt, the most-well known linear regression model.  Despite its wide applicability, it often gives undesireable results when the data deviate from its underlying normal model.  In particular, it is quite sensitive to outliers in the data.  In this post, we will illustrate this sensitivity and then show that changing the error distribution results in a more robust regression model.\n\nWe will use one of the data sets from [Anscombe's quartet](http://en.wikipedia.org/wiki/Anscombe%27s_quartet) to illustrate these concepts.  Anscombe's quartet is a well-known group of four data sets that illustrates the importance of exploratory data analysis 

AustinRochford

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$\\int_a^b f'(t) dt = f(b) - f(a)$"
   ]
  },
  {

AustinRochford

AustinRochford

AustinRochford