5 - Priors Over Grouped Variables.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:055e821907f6845906c0702b728e0cf174b49f7e0d21af713fefae9d599047ef"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "[Demographic Forecasting](http://gking.harvard.edu/files/abs/smooth-Abs.shtml) - [Chapter 5: Priors Over Grouped Variables](http://gking.harvard.edu/files/gking/files/chapter5_1.pdf)"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Model"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$\n",
      "m_{at} \\sim \\mathcal{N} \\bigg( \\mu_{at}, \\frac{{\\sigma_{a}}^2} {b_{at}} \\bigg)\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$\n",
      "\\mu_{at} = \\beta_{a} \\textbf{Z}_{at}\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Joint Likelihood ([Eq 4.1](http://gking.harvard.edu/files/gking/files/chapter4_1.pdf))"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$ \n",
      "P \\big(\\beta, \\theta, \\sigma \\mid m \\big) \\propto \n",
      "P \\big(m \\mid \\beta, \\sigma \\big) \\cdot \\big[ P \\big(\\beta \\mid \\theta \\big) \\cdot P \\big(\\theta \\big) \\cdot P \\big(\\sigma \\big) \\big]\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Data Likelihood ([Eq 9.3](http://gking.harvard.edu/files/gking/files/chapter9_1.pdf))"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$ \n",
      "P \\big(m \\mid \\beta, \\sigma \\big) \\propto \n",
      "\\bigg(\\prod_{a} {(\\sigma_{a}^{-2})} ^ {\\frac{T}{2}} \\bigg) \\cdot \n",
      "\\Bigg(-\\frac{1}{2} \\sum_{a} \\bigg( \\frac{1}{\\sigma_{a}^{2}} \\sum_{t} {\\big(m_{at} - \\beta_{a} \\textbf{Z}_{at} \\big)}^{2} \\bigg) \\Bigg) \n",
      "$$"
     ]
    },
    {
     "cell_type": "heading",
     "level": 4,
     "metadata": {},
     "source": [
      "Priors"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Prior on data variance ([Eq 9.6](http://gking.harvard.edu/files/gking/files/chapter9_1.pdf))"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$\n",
      "\\frac{1}{\\sigma_{i}^{2}} \\sim \\mathcal{G} \\Big(\\frac{\\mathbb{d}}{2}, \\frac{\\mathbb{e}}{2} \\Big)\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Prior on coefficients ([Eq 5.16](http://gking.harvard.edu/files/gking/files/chapter5_1.pdf))"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$ \n",
      "P \\big( \\beta \\mid \\theta \\big) \\propto\n",
      "\\exp \\bigg( - \\frac{1}{2} \\theta \n",
      "\\sum_{a a^{\\prime}} W_{a a^{\\prime}}^{\\textrm{age},\\mathbb{n}}\n",
      "{\\beta_{a}}^{\\prime} \\textbf{C}_{a a^{\\prime}} \\beta_{a^{\\prime}} \\bigg)\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$ \n",
      "W_{a a^{\\prime}}^{\\textrm{age}, \\mathbb{n}} = \n",
      "A^{-1} \\big(D^{\\textrm{age},\\mathbb{n}} \\big)^{\\prime} D^{\\textrm{age},\\mathbb{n}}\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$\n",
      "\\sum_{a^{\\prime}} D^{\\textrm{age},\\mathbb{n}} \\mu_{a^{\\prime} t} =\n",
      "\\frac{d^{\\mathbb{n}} \\mu(a,t)} {da^{\\mathbb{n}}}\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$\n",
      "\\textbf{C}_{a a^{\\prime}} =\n",
      "\\frac{1}{T}\n",
      "{\\textbf{Z}_{a}}^{\\prime} \\textbf{Z}_{a^{\\prime}}\n",
      "$$"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Prior on smoothing parameter ([Eq 9.10](http://gking.harvard.edu/files/gking/files/chapter9_1.pdf))"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "$$\n",
      "\\theta \\sim \\mathcal{G} \\Big(\\frac{\\mathbb{f}}{2}, \\frac{\\mathbb{g}}{2} \\Big)\n",
      "$$"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "import seaborn as sns\n",
      "import matplotlib.pyplot as plt\n",
      "import numpy as np\n",
      "try:\n",
      "    import fbd\n",
      "    import PyMB\n",
      "    from PyMB import magic\n",
      "except:\n",
      "    import getpass, sys\n",
      "    sys.path.append('/homes/{user}/'.format(user=getpass.getuser())) # your directory here...\n",
      "    import fbd\n",
      "    import PyMB\n",
      "    from PyMB import magic"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "/usr/bin/anaconda/lib/python2.7/site-packages/matplotlib/__init__.py:1312: UserWarning:  This call to matplotlib.use() has no effect\n",
        "because the backend has already been chosen;\n",
        "matplotlib.use() must be called *before* pylab, matplotlib.pyplot,\n",
        "or matplotlib.backends is imported for the first time.\n",
        "\n",
        "  warnings.warn(_use_error_msg)\n"
       ]
      },
      {
       "javascript": [
        "IPython.config.cell_magic_highlight['magic_clike'] = {'reg':[/^%%PyMB/]};"
       ],
       "metadata": {},
       "output_type": "display_data"
      }
     ],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%PyMB MAGiK -NO_WRAP\n",
      "// UTILITY FUNCTIONS\n",
      "\n",
      "    // NAME:    first_diff\n",
      "    // DESC:    returns a matrix which can be used to compute the vector of first differences of another vector.\n",
      "    // SOURCE:  https://github.com/IQSS/YourCast/blob/396408b87c032bc7588f28c75d3ea9801f8e1b93/R/make.priors.R#L604\n",
      "    // IN:      d, scalar, dimension of the vector for which we want to compute the first differences\n",
      "    // OUT:     (d-1) x d matrix of the form:\n",
      "    //             -1    1    0    0    0  ...\n",
      "    //              0   -1    1    0    0  ...\n",
      "    //              0    0   -1    1    0  ...\n",
      "    //              0    0    0   -1    1  ...\n",
      "    //             ...  ...  ...  ...  ... ...\n",
      "    template<class Type>\n",
      "    matrix<Type> first_diff(int d) { \n",
      "        matrix<Type> M(d-1,d);\n",
      "        for (size_t i = 0; i < d-1; i++) {\n",
      "            for (size_t j = 0; j < d; j++) {\n",
      "                // -1 on the diagonal\n",
      "                if (i == j) M(i,j) = -1.;\n",
      "                // +1 on the upper off-diagonal\n",
      "                else if (i+1 == j) M(i,j) = 1.;\n",
      "                // 0 everywhere else\n",
      "                else M(i,j) = 0.;\n",
      "            }\n",
      "        }\n",
      "        return M;\n",
      "    }\n",
      "    \n",
      "    // NAME:    n_diff\n",
      "    // DESC:    returns a matrix which can be used to compute the vector of n-th differences of another vector.\n",
      "    // SOURCE:  https://github.com/IQSS/YourCast/blob/396408b87c032bc7588f28c75d3ea9801f8e1b93/R/make.priors.R#L658\n",
      "    // IN:      n, scalar\n",
      "    //          d, scalar, dimension of the vector for which we want to compute the first differences\n",
      "    // OUT:     (d-n) x d matrix. It is simply the matrix obtained by iterating the first difference operator n times. \n",
      "    //          It resembles the discretization of the $n$-th derivative and can be used to define smoothness functionals.\n",
      "    template<class Type>\n",
      "    matrix<Type> n_diff(int n, int d) { \n",
      "        matrix<Type> M = first_diff<Type>(d);\n",
      "        // multiply first differences for each higher order\n",
      "        if (n > 1) \n",
      "            for (size_t i = 1; i < n; i++)\n",
      "                M = first_diff<Type>(M.rows()) * M;\n",
      "        return M;\n",
      "    }\n",
      "        \n",
      "    // NAME:    deriv\n",
      "    // DESC:    returns the n-th difference matrix with rows added to ensure conformability when n > 1\n",
      "    // SOURCE:  https://github.com/IQSS/YourCast/blob/396408b87c032bc7588f28c75d3ea9801f8e1b93/R/make.priors.R#L740\n",
      "    // IN:      n, scalar\n",
      "    //          d, scalar, dimension of the vector for which we want to compute the first differences\n",
      "    // OUT:     d x d matrix. It is simply the matrix obtained by iterating the first difference operator n times. \n",
      "    //          First/last rows are repeated to conform to d x d when n > 1 \n",
      "    template<class Type>\n",
      "    matrix<Type> deriv(int n, int d) { \n",
      "\n",
      "        matrix<Type> M(d,d);\n",
      "        M.setZero();\n",
      "\n",
      "        // simply return I if n==0\n",
      "        if (n == 0) {\n",
      "            for (size_t i = 0; i < d; i++)\n",
      "                M(i,i) = 1.;\n",
      "        }\n",
      "\n",
      "        // otherwise find n_diff and add requisite rows to make it conformable\n",
      "        else {\n",
      "\n",
      "            // find n_diff\n",
      "            matrix<Type> D = n_diff<Type>(n, d);\n",
      "\n",
      "            // for n == 1, repeat the first row\n",
      "            if (n == 1)\n",
      "                M.row(0) = D.row(0);   \n",
      "\n",
      "            // for n >= 2, copy the first and last rows in n/2 times each\n",
      "            if (n >= 2) {\n",
      "                for (size_t i = 0; i < n/2; i++) {\n",
      "                    M.row(i) = D.row(0);\n",
      "                    M.row(d-(i+1)) = D.row(D.rows()-1);\n",
      "                } \n",
      "                // for odd n, repeat the first row\n",
      "                if (n % 2 == 1) M.row(ceil(n/2)) = D.row(0);\n",
      "            }\n",
      "\n",
      "            // fill in the remaining rows from D\n",
      "            for (size_t i = 0; i < D.rows(); i++)\n",
      "                M.row(ceil(n/2) + i + (n % 2)) = D.row(i);\n",
      "        }\n",
      "        return M;\n",
      "    }\n",
      "        \n",
      "    // NAME:    derivative_prior\n",
      "    // DESC:    returns the matrix W that defines a mixed smoothness prior\n",
      "    // SOURCE:  https://github.com/IQSS/YourCast/blob/396408b87c032bc7588f28c75d3ea9801f8e1b93/R/make.priors.R#L763\n",
      "    // IN:      d: scalar\n",
      "    //              dimension of the desired smoothing matrix\n",
      "    //          der_v: vector\n",
      "    //              weight of each derivative in the mixed smoothness function, \n",
      "    //              starting from the derivative of order 0 (the identity operator).\n",
      "    //                  Example 1: der_v = c(0,0,1) corresponds to a smoothness functional which\n",
      "    //                      penalizes the 2nd derivative (0*identity + 0*1st derivative + 1*2nd derivative)\n",
      "    //                  Example 2: der_v = c(0,1,1) corresponds to a smoothness functional\n",
      "    //                      which penalizes equally the 1st and 2nd derivative \n",
      "    //                      (0*identity + 1*1st derivative + 1*2nd derivative)\n",
      "    // OUT:     W: d x d matrix, the matrix which defines the smoothness prior.\n",
      "    template<class Type>\n",
      "    matrix<Type> derivative_prior(int d, vector<Type> der_v) {\n",
      "        // start with zeros\n",
      "        matrix<Type> W(d,d);\n",
      "        W.setZero();\n",
      "\n",
      "        // find weights (see YourCast for possible extensions - for now, just weight everything equally)\n",
      "        // should be diagonal 1/d\n",
      "        matrix<Type> weights(d,d);\\\n",
      "        weights.setZero();\n",
      "        for (size_t i = 0; i < d; i++) \n",
      "            weights(i,i) = 1. / d;\n",
      "\n",
      "        // add on the weights for each derivative included in the prior\n",
      "        for (size_t i = 0; i < der_v.size(); i++) {\n",
      "            if (der_v(i) != 0.) {\n",
      "                matrix<Type> m = deriv<Type>(i, d);\n",
      "                W = W + der_v(i) * (m.transpose() * weights * m);\n",
      "            }\n",
      "        }\n",
      "        return W;\n",
      "    }\n",
      "\n",
      "    // NAME:    mul_vec_mat_vec\n",
      "    // DESC:    returns t(A) * B * C\n",
      "    //          most commonly used to calculate e.g. B_{a}' * C * B_{a'}\n",
      "    //          removes NaNs (by treating them as zero)\n",
      "    // IN:      A: vector of length n\n",
      "    //          B: matrix of size n*n\n",
      "    //          C: vector of length n\n",
      "    // OUT:     scalar <Type>\n",
      "    template<class Type>\n",
      "    Type mul_vec_mat_vec(vector<Type> A, matrix<Type> B, vector<Type> C) {\n",
      "        for (size_t i = 0; i < A.rows(); i++) {\n",
      "            if (CppAD::isnan(A(i))) A(i) = 0;\n",
      "        }\n",
      "        for (size_t i = 0; i < B.rows(); i++) {\n",
      "            for (size_t j = 0; j < B.cols(); j++) {\n",
      "                if (CppAD::isnan(B(i,j))) B(i,j) = 0;\n",
      "            }\n",
      "        }\n",
      "        for (size_t i = 0; i < C.rows(); i++) {\n",
      "            if (CppAD::isnan(C(i))) C(i) = 0;\n",
      "        }\n",
      "        return sum(vector<Type>(A * vector<Type>(B * C)));\n",
      "    }\n",
      "\n",
      "    // NAME:    mean_skip_NA\n",
      "    // DESC:    calculates the mean of a vector while excluding NAs\n",
      "    // IN:      vector<Type>\n",
      "    // OUT:     scalar<Type>\n",
      "    template<class Type>\n",
      "    Type mean_skip_NA(vector<Type> v) {\n",
      "        int n = 0;\n",
      "        Type s = 0.;\n",
      "        for (size_t i = 0; i < v.rows(); i++) {\n",
      "            if (!CppAD::isnan(v(i))) {\n",
      "                n += 1;\n",
      "                s += v(i);\n",
      "            }\n",
      "        }\n",
      "        return (s / n);\n",
      "    }\n",
      "\n",
      "    // NAME:    row_mean\n",
      "    // DESC:    calculates the row means of a matrix while excluding NAs\n",
      "    // IN:      matrix<Type>\n",
      "    // OUT:     vector<Type>\n",
      "    template<class Type>\n",
      "    vector<Type> row_mean(matrix<Type> m) {\n",
      "        vector<Type> means(m.rows());\n",
      "        for (size_t i = 0; i < m.rows(); i++) {\n",
      "            means(i) = mean_skip_NA(vector<Type>(m.row(i)));\n",
      "        }\n",
      "        return means;\n",
      "    }\n",
      "\n",
      "// MODEL OBJECTIVE FUNCTION\n",
      "    template<class Type>\n",
      "    Type objective_function<Type>::operator()(){\n",
      "    // DATA\n",
      "        // dimensions\n",
      "            DATA_INTEGER(A);    // # of ages\n",
      "            DATA_INTEGER(T);    // # of years\n",
      "            DATA_INTEGER(k);    // # of covariates\n",
      "        // covariate matrix (rows: A,k; cols: T)\n",
      "            DATA_MATRIX(Z);\n",
      "        // data matrix (rows: A; cols: T)\n",
      "            DATA_MATRIX(m);\n",
      "    \n",
      "    // HYPERPARAMETERS\n",
      "        // order of smoothness for age prior\n",
      "        // e.g. (0,0,1) would be smoothness on second-order derivative\n",
      "        //      (0,1,1) would be equal weights on 1st and 2nd\n",
      "            DATA_VECTOR(age_deriv_n);   \n",
      "        // 1/sigma^2 ~ gamma(a/2, b/2)\n",
      "            DATA_SCALAR(sigma_a);\n",
      "            DATA_SCALAR(sigma_b);\n",
      "        // theta ~ gamma(a/2, b/2)\n",
      "            DATA_SCALAR(theta_a);\n",
      "            DATA_SCALAR(theta_b);\n",
      "        \n",
      "    // RESCALE DATA\n",
      "        // need to remove the mean from each cross sectional unit so that our priors are on the smoothness of the deviation\n",
      "            vector<Type> m_means(A);\n",
      "            m_means = row_mean(m);\n",
      "            matrix<Type> m_adj(A,T);\n",
      "            for (size_t a = 0; a < A; a++) {\n",
      "                for (size_t t = 0; t < T; t++) {\n",
      "                    m_adj(a,t) = m(a,t) - m_means(a);\n",
      "                }\n",
      "            }\n",
      "            REPORT(m_adj);\n",
      "\n",
      "    // PARAMETERS\n",
      "        // betas (rows: A; cols: k)\n",
      "            PARAMETER_MATRIX(Beta);\n",
      "        // log(sigma^2) (ages)\n",
      "            PARAMETER_VECTOR(log_sigma_2);\n",
      "        // log(theta^2)\n",
      "            PARAMETER(log_theta_2);\n",
      "    \n",
      "    // CACHED TRANSFORMATIONS\n",
      "        // W matrix of age weights\n",
      "            matrix<Type> W = derivative_prior<Type>(A, age_deriv_n);\n",
      "        // C = ZZ'*(1/T) by age pairs\n",
      "            matrix<Type> C[A][A];\n",
      "            for (size_t a1 = 0; a1 < A; a1++) {\n",
      "                for (size_t a2 = 0; a2 < A; a2++) {\n",
      "                    C[a1][a2] = Eigen::operator*((1. / T), (Z.block(a1*k, 0, k, T) * Z.block(a2*k, 0, k, T).transpose()));\n",
      "                }\n",
      "            }\n",
      "    \n",
      "    // BEGIN LIKELIHOOD\n",
      "        parallel_accumulator<Type> nll(this);\n",
      "    \n",
      "    // PRIORS\n",
      "        // p(theta)\n",
      "            Type theta = exp(0.5 * log_theta_2);\n",
      "            nll -= dgamma(theta, theta_a/2., theta_b/2., true);\n",
      "        // p(sigma)\n",
      "            vector<Type> sigma_2(A);\n",
      "            for (size_t a = 0; a < A; a++) {\n",
      "                sigma_2(a) = exp(log_sigma_2(a));\n",
      "                nll -= dgamma(1./sigma_2(a), sigma_a/2., sigma_b/2., true);\n",
      "            }\n",
      "        // p(Beta|theta)\n",
      "            // extract vectors of Beta by age\n",
      "                vector<Type> Beta_vec[A];\n",
      "                for (size_t a = 0; a < A; a++) {\n",
      "                    Beta_vec[a] = vector<Type>(Beta.row(a));\n",
      "                }\n",
      "            // calculate the sum of WB'CB\n",
      "                Type sum_WBCB = 0;\n",
      "                for (size_t a1 = 0; a1 < A; a1++) {\n",
      "                    for (size_t a2 = 0; a2 < A; a2++) {\n",
      "                        sum_WBCB += W(a1,a2) * mul_vec_mat_vec(Beta_vec[a1], C[a1][a2], Beta_vec[a2]);\n",
      "                    }\n",
      "                }\n",
      "            // add to the likelihood\n",
      "                nll -= (-1./2.) * theta * sum_WBCB;\n",
      "               \n",
      "    // DATA LIKELIHOOD\n",
      "        matrix<Type> mu(A,T);\n",
      "        matrix<Type> BZ(A,T);\n",
      "        for (size_t a = 0; a < A; a++) {\n",
      "            for (size_t t = 0; t < T; t++) {\n",
      "                BZ(a,t) = Type(0.);\n",
      "                for (size_t kk = 0; kk < k; kk++) {\n",
      "                    BZ(a,t) += Z(a*k + kk, t) * Beta(a,kk);\n",
      "                }\n",
      "                mu(a,t) = m_means(a) + BZ(a,t);\n",
      "                if (!CppAD::isnan(BZ(a,t)) & !CppAD::isnan(m_adj(a,t))) {\n",
      "                    nll -= dnorm(m_adj(a,t), BZ(a,t), sqrt(sigma_2(a)), true);\n",
      "                }\n",
      "            }\n",
      "        }\n",
      "        REPORT(mu);\n",
      "    \n",
      "    // RETURN LIKELIHOOD\n",
      "        return nll;\n",
      "    }"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Created model MAGiK.\n",
        "Saving model to tmb_tmp/MAGiK.cpp.\n",
        "Compiled in 34.7s.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n"
       ]
      }
     ],
     "prompt_number": 2
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Data"
     ]
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Load"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# if you need to remake the data:\n",
      "# %run \"0 - Shared Data & Functions.ipynb\"\n",
      "alldata = pd.read_csv('data/lee_carter.csv')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Males, all-cause USA mortality"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = alldata.ix[(alldata.iso3 == 'USA') & (alldata.sex == 1)]\n",
      "data.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>LDI</th>\n",
        "      <th>age</th>\n",
        "      <th>educ</th>\n",
        "      <th>iso3</th>\n",
        "      <th>log_risk_subtracted_mx</th>\n",
        "      <th>log_total_mx</th>\n",
        "      <th>pop</th>\n",
        "      <th>region</th>\n",
        "      <th>risk_subtracted_mx</th>\n",
        "      <th>sex</th>\n",
        "      <th>super_region</th>\n",
        "      <th>total_mx</th>\n",
        "      <th>year</th>\n",
        "      <th>ln_LDI</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>5712</th>\n",
        "      <td> 24227.80</td>\n",
        "      <td> 0</td>\n",
        "      <td> 12.52580</td>\n",
        "      <td> USA</td>\n",
        "      <td>-0.794426</td>\n",
        "      <td>-0.791471</td>\n",
        "      <td> 34086.64314</td>\n",
        "      <td> 10</td>\n",
        "      <td> 0.451841</td>\n",
        "      <td> 1</td>\n",
        "      <td> 2</td>\n",
        "      <td> 0.453178</td>\n",
        "      <td> 1980</td>\n",
        "      <td> 10.095256</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5714</th>\n",
        "      <td> 24639.93</td>\n",
        "      <td> 0</td>\n",
        "      <td> 12.58925</td>\n",
        "      <td> USA</td>\n",
        "      <td>-0.868012</td>\n",
        "      <td>-0.865057</td>\n",
        "      <td> 34745.42902</td>\n",
        "      <td> 10</td>\n",
        "      <td> 0.419785</td>\n",
        "      <td> 1</td>\n",
        "      <td> 2</td>\n",
        "      <td> 0.421027</td>\n",
        "      <td> 1981</td>\n",
        "      <td> 10.112124</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5716</th>\n",
        "      <td> 24836.26</td>\n",
        "      <td> 0</td>\n",
        "      <td> 12.65018</td>\n",
        "      <td> USA</td>\n",
        "      <td>-0.916482</td>\n",
        "      <td>-0.913527</td>\n",
        "      <td> 35431.22417</td>\n",
        "      <td> 10</td>\n",
        "      <td> 0.399923</td>\n",
        "      <td> 1</td>\n",
        "      <td> 2</td>\n",
        "      <td> 0.401107</td>\n",
        "      <td> 1982</td>\n",
        "      <td> 10.120060</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5718</th>\n",
        "      <td> 25145.43</td>\n",
        "      <td> 0</td>\n",
        "      <td> 12.70797</td>\n",
        "      <td> USA</td>\n",
        "      <td>-0.988800</td>\n",
        "      <td>-0.985845</td>\n",
        "      <td> 36107.21003</td>\n",
        "      <td> 10</td>\n",
        "      <td> 0.372023</td>\n",
        "      <td> 1</td>\n",
        "      <td> 2</td>\n",
        "      <td> 0.373124</td>\n",
        "      <td> 1983</td>\n",
        "      <td> 10.132431</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5720</th>\n",
        "      <td> 25701.04</td>\n",
        "      <td> 0</td>\n",
        "      <td> 12.76154</td>\n",
        "      <td> USA</td>\n",
        "      <td>-1.052350</td>\n",
        "      <td>-1.049396</td>\n",
        "      <td> 36746.81840</td>\n",
        "      <td> 10</td>\n",
        "      <td> 0.349116</td>\n",
        "      <td> 1</td>\n",
        "      <td> 2</td>\n",
        "      <td> 0.350149</td>\n",
        "      <td> 1984</td>\n",
        "      <td> 10.154287</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 4,
       "text": [
        "           LDI  age      educ iso3  log_risk_subtracted_mx  log_total_mx  \\\n",
        "5712  24227.80    0  12.52580  USA               -0.794426     -0.791471   \n",
        "5714  24639.93    0  12.58925  USA               -0.868012     -0.865057   \n",
        "5716  24836.26    0  12.65018  USA               -0.916482     -0.913527   \n",
        "5718  25145.43    0  12.70797  USA               -0.988800     -0.985845   \n",
        "5720  25701.04    0  12.76154  USA               -1.052350     -1.049396   \n",
        "\n",
        "              pop  region  risk_subtracted_mx  sex  super_region  total_mx  \\\n",
        "5712  34086.64314      10            0.451841    1             2  0.453178   \n",
        "5714  34745.42902      10            0.419785    1             2  0.421027   \n",
        "5716  35431.22417      10            0.399923    1             2  0.401107   \n",
        "5718  36107.21003      10            0.372023    1             2  0.373124   \n",
        "5720  36746.81840      10            0.349116    1             2  0.350149   \n",
        "\n",
        "      year     ln_LDI  \n",
        "5712  1980  10.095256  \n",
        "5714  1981  10.112124  \n",
        "5716  1982  10.120060  \n",
        "5718  1983  10.132431  \n",
        "5720  1984  10.154287  "
       ]
      }
     ],
     "prompt_number": 4
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Mark out of sample 2006+"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data['m_obs'] = data['log_total_mx']\n",
      "data['insample'] = data.year <= 2005\n",
      "data['m_train'] = data.ix[data.insample, 'm_obs']"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 5
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Drop after 2013"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = data.ix[data.year <= 2013]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 6
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Convert to panel"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = data.set_index(['age','year']).to_panel()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 7
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Select covariates"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "covariates = ['ln_LDI', 'educ']\n",
      "k = len(covariates)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 8
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Format data for model"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# matrix of covariates (rows: A,k, cols: T)\n",
      "Z = np.hstack([data[c].as_matrix() for c in covariates]).reshape([data.shape[1]*k, data.shape[2]], order='C')\n",
      "# matrix of in-sample log rates\n",
      "m = data['m_train'].as_matrix()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 9
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Select priors"
     ]
    },
    {
     "cell_type": "heading",
     "level": 4,
     "metadata": {},
     "source": [
      "Priors on $\\theta$ and $\\sigma$"
     ]
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "To convert to $\\mathcal{G}(\\mathbb{a},\\mathbb{b})$ parameters"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# function to define a gamma prior on something given its mean and sd\n",
      "# source: https://github.com/IQSS/YourCast/blob/955a88043fa97d71b922585b5bcd28cc5738d75c/R/compute.sigma.R#L194\n",
      "def sigma_param(m, std):\n",
      "    from scipy.special import gammaln\n",
      "    from scipy.optimize import brentq\n",
      "    def fun(x,m,v):\n",
      "        return np.sqrt((x/2.)-1.) * np.exp(gammaln((x/2.)-0.5) - gammaln(x/2.)) - (m / np.sqrt(m**2. + v))\n",
      "    v = std**2.\n",
      "    aux = (m**2.) / (v + m**2.)\n",
      "    d = brentq(fun, 2.0000001, 1000, args=(m,v))\n",
      "    e = (d - 2.) * (v + m**2.)\n",
      "    return [d,e]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 10
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Empirical prior on $\\sigma$"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def make_sigma_priors(m):\n",
      "    # keep just insample data\n",
      "    m_in = m[np.isnan(m) == False]\n",
      "    m_in = m_in.reshape([m.shape[0], m_in.shape[0]/m.shape[0]])\n",
      "    # adjust it by subtracting off row means\n",
      "    m_means = m_in.mean(axis=1)\n",
      "    m_adj = m_in - np.repeat(m_means, m_in.shape[1]).reshape([m_means.shape[0],m_in.shape[1]])\n",
      "    # find sd by age\n",
      "    sigma = m_adj.std(axis=1)\n",
      "    # find average sd\n",
      "    mean_sigma = np.mean(sigma)\n",
      "    # and sd of sd\n",
      "    sd_sigma = np.std(sigma)\n",
      "    # convert to gamma priors\n",
      "    return sigma_param(mean_sigma, sd_sigma)\n",
      "sigma_priors = make_sigma_priors(m)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 11
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Find priors for $\\theta$"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def make_theta_priors(m):\n",
      "    # keep just insample data\n",
      "    m_in = m[np.isnan(m) == False]\n",
      "    m_in = m_in.reshape([m.shape[0], m_in.shape[0]/m.shape[0]])\n",
      "    # adjust it by subtracting off row means\n",
      "    m_means = m_in.mean(axis=1)\n",
      "    m_adj = m_in - np.repeat(m_means, m_in.shape[1]).reshape([m_means.shape[0],m_in.shape[1]])\n",
      "    # find the sqrt of the differences between age groups\n",
      "    m_diff = np.sqrt(np.abs(np.diff(m_adj, axis=0)))\n",
      "    # use mean and sd of that difference\n",
      "    mn = np.mean(m_diff)\n",
      "    sd = np.std(m_diff)\n",
      "    return sigma_param(mn, sd)\n",
      "theta_priors = make_theta_priors(m)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "MAGiK.data = {\n",
      "    'A':    m.shape[0],\n",
      "    'T':    m.shape[1],\n",
      "    'k':    k,\n",
      "    'Z':    Z,\n",
      "    'm':    m,\n",
      "    'sigma_a':  sigma_priors[0], \n",
      "    'sigma_b':  sigma_priors[1],\n",
      "    'theta_a':  theta_priors[0],\n",
      "    'theta_b':  theta_priors[1],\n",
      "    'age_deriv_n': 1.\n",
      "}"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 13
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Model"
     ]
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Initial values"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "MAGiK.init = {\n",
      "    'log_sigma_2':  np.zeros(m.shape[0]),\n",
      "    'log_theta_2':  0.,\n",
      "    'Beta':         np.zeros([m.shape[0], k]),\n",
      "}"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 14
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Random effects"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "MAGiK.random = ['Beta']\n",
      "#MAGiK.random = []"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "heading",
     "level": 5,
     "metadata": {},
     "source": [
      "Fit"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "MAGiK.optimize()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Order of parameters:\n",
        "[1] \"Beta\"        \"log_theta_2\" \"log_sigma_2\"\n",
        "Not matching template order:\n",
        "[1] \"Beta\"        \"log_sigma_2\" \"log_theta_2\"\n",
        "Your parameter list has been re-ordered.\n",
        "(Disable this warning with checkParameterOrder=FALSE)\n",
        "Matching hessian patterns... "
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Done\n",
        "\n",
        "Model optimization complete in 0.3s.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "\n",
        "--------------------------------------------------------------------------------\n",
        "\n",
        "\n",
        "Simulated 100 draws in 0.2s.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "Beta:\n",
        "\tmean\t[ 4.09843489  2.79952378 ..., -0.15752273 -0.07780603]\n",
        "\tsd\t[ 4.33446514  4.32492733 ...,  0.34832051  0.34829444]\n",
        "\tdraws\t[[[ 1.85963106  8.81333641 ...,  5.49467772  0.3127279 ],\n",
        "\t\t ...,\n",
        "\t\t [-0.83426137  0.25765323 ..., -0.02499556  0.73637767]]]\n",
        "\tshape\t(20, 2, 100)\n"
       ]
      }
     ],
     "prompt_number": 16
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data['m_pred'] = MAGiK.report('mu')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 17
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def age_time_plots(panel, y_obs='y_obs', y_mean='y_mean', y_lower='y_lower', y_upper='y_upper', title='', y_label='$\\mathrm{ln}(m_{x})$'):\n",
      "    ## setup figure\n",
      "    f, axs = plt.subplots(nrows=1, ncols=2, figsize=(15,7), gridspec_kw={'width_ratios':[1,1.2]})\n",
      "    f.suptitle(title, fontsize=15)\n",
      "    axs[0].set_ylabel(y_label, fontsize=15)\n",
      "    first_year = panel.axes[2][0]\n",
      "    last_year = panel.axes[2][-1]\n",
      "\n",
      "    ## time plot\n",
      "    ax = axs[0]\n",
      "    # label x\n",
      "    ax.set_xlabel('Year', fontsize=12)\n",
      "    ax.set_xlim(left=first_year-0.5, right=last_year+2)\n",
      "    # create palette\n",
      "    pal = sns.hls_palette(n_colors=len(panel.axes[1]), l=0.6)\n",
      "    # add scatters and lines\n",
      "    sns.set_context(rc={'lines.markeredgewidth': 0.1}) # bugfix for seaborn making markers invisible\n",
      "    for i,a in enumerate(panel.axes[1]):\n",
      "        if y_lower in panel.items and y_upper in panel.items:\n",
      "            ax.fill_between(panel.axes[2], panel[y_lower].ix[a], panel[y_upper].ix[a], color=pal[i], alpha=0.6)\n",
      "        ax.plot(panel.axes[2], panel[y_obs].ix[a], 'o', markerfacecolor=pal[i], markersize=5, alpha=0.9)\n",
      "        ax.plot(panel.axes[2], panel[y_mean].ix[a], '-', color=pal[i])\n",
      "        ax.text(last_year+1, panel[y_mean].ix[a,last_year], a, verticalalignment='center', color=pal[i], alpha=0.9)\n",
      "\n",
      "    ## age plot\n",
      "    ax = axs[1]\n",
      "    # label x\n",
      "    ax.set_xlabel('Age', fontsize=12)\n",
      "    ax.set_xlim(left=panel.axes[1][0], right=panel.axes[1][-1])\n",
      "    # create palette\n",
      "    pal = sns.hls_palette(n_colors=len(panel.axes[2]), l=0.6)\n",
      "    # add lines\n",
      "    for i,y in enumerate(panel.axes[2]):\n",
      "        ax.plot(panel[y_mean].index, panel[y_mean][y], c=pal[i], alpha=0.9)\n",
      "    \n",
      "    ## colorbar for age plot (see http://stackoverflow.com/questions/8342549/matplotlib-add-colorbar-to-a-sequence-of-line-plots)\n",
      "    ctb = matplotlib.colors.LinearSegmentedColormap.from_list('custombar', pal, N=2048)\n",
      "    sm = plt.cm.ScalarMappable(cmap=ctb, norm=plt.Normalize(vmin=first_year, vmax=last_year))\n",
      "    sm._A = []\n",
      "    # Set colorbar, aspect ratio\n",
      "    cbar = plt.colorbar(sm, alpha=0.9, aspect=16, shrink=0.6)\n",
      "    cbar.solids.set_edgecolor('face')\n",
      "    # Remove colorbar container frame\n",
      "    cbar.outline.set_visible(False)\n",
      "    # Fontsize for colorbar ticklabels\n",
      "    cbar.ax.tick_params(labelsize=16)\n",
      "    # Customize colorbar tick labels\n",
      "    mytks = np.arange(first_year,last_year+1)[np.arange(first_year,last_year+1) % 10 == 0]\n",
      "    cbar.set_ticks(mytks)\n",
      "    cbar.ax.set_yticklabels([str(a) for a in mytks], fontsize=8)\n",
      "    # Remove color bar tick lines, while keeping the tick labels\n",
      "    cbarytks = plt.getp(cbar.ax.axes, 'yticklines')\n",
      "    plt.setp(cbarytks, visible=False)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 18
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "age_time_plots(data, y_obs='m_obs', y_mean='m_pred')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAA2sAAAG9CAYAAACLTe6LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+j/pDvQ1kN9hfQAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7ff5e6856950>"
       ]
      }
     ],
     "prompt_number": 19
    }
   ],
   "metadata": {}
  }
 ]
}