{
 "metadata": {
  "name": "",
  "signature": "sha256:61cd9a7b8c09b2ce13199265ad30b86550646863aa2e3d2eab6afa4cc3965fc4"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "# Live oaks introgression manuscript\n",
      "## Notebook 5: dadi analysis \n",
      "\n",
      "#### D. Eaton, A. Hipp, A. Gonzalez-Rodriguez & J. Cavender-Bares\n",
      "##### contact: deren.eaton@yale.edu  \n",
      "\n",
      "-----------------------   "
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## Model selection and parameter estimation in _dadi_\n",
      "\n",
      "From the tutorial: \"\u2202a\u2202i is a powerful software tool for simulating the joint frequency spectrum (FS) of genetic variation among multiple populations and employing the FS for population-genetic inference\"  \n",
      "\n",
      "Here I will use the _dadi_ Python package to analyze the SFS from RADseq data assembled in _pyRAD_ for three oak species. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## import libraries\n",
      "import dadi\n",
      "import numpy as np\n",
      "import itertools as itt\n",
      "import pandas as pd\n",
      "import sys, os\n",
      "\n",
      "## prettier printing of arrays\n",
      "pd.set_option('precision',4)\n",
      "pd.set_option('chop_threshold', 0.001)\n",
      "\n",
      "## plotting libraries\n",
      "import matplotlib.pyplot as plt\n",
      "import seaborn as sns\n",
      "%matplotlib inline"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Convert SNP data to FS   \n",
      "For large data sets fs will be counted from many SNPs and so is easiest to read in from a file. In this case I convert a `.loci` output file with thousands of RADseq loci from an analysis in _pyRAD_ into the SNP format used by _dadi_, and then extract the site frequency spectrum from that file. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## a function to select most common element in a list\n",
      "def most_common(L):\n",
      "    return max(itt.groupby(sorted(L)),\n",
      "               key=lambda(x, v):(len(list(v)),\n",
      "               -L.index(x)))[0]\n",
      "\n",
      "## a function to get alleles from ambiguous bases\n",
      "def unstruct(amb):\n",
      "    \" returns bases from ambiguity code\"\n",
      "    D = {\"R\":[\"G\",\"A\"],\n",
      "         \"K\":[\"G\",\"T\"],\n",
      "         \"S\":[\"G\",\"C\"],\n",
      "         \"Y\":[\"T\",\"C\"],\n",
      "         \"W\":[\"T\",\"A\"],\n",
      "         \"M\":[\"C\",\"A\"]}\n",
      "    if amb in D:\n",
      "        return D.get(amb)\n",
      "    else:\n",
      "        return [amb,amb]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## parse the loci\n",
      "locifile = open(\"analysis_pyrad/outfiles/virentes_c85d6m4p5.loci\")\n",
      "loci = locifile.read().strip().split(\"|\")"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## define focal populations and the outgroups\n",
      "cuba        = ['CUCA4','CUVN10','CUSV6']\n",
      "centralA    = ['CRL0030','CRL0001','HNDA09','BZBB1','MXSA3017']\n",
      "virginiana  = ['FLSF33','FLBA140','SCCU3','LALC2'] \n",
      "geminata    = [\"FLCK18\",\"FLSF54\",\"FLWO6\",\"FLAB109\"]\n",
      "minima      = [\"FLSF47\",\"FLMO62\",\"FLSA185\",\"FLCK216\"]\n",
      "outgroup    = [\"DO\",\"DU\",\"EN\",\"AR\",\"CH\",\"NI\",\"HE\",\"BJSB3\",\"BJVL19\",\"BJSL25\"]\n",
      "\n",
      "## florida is a composite of three 'species'\n",
      "florida = virginiana+geminata+minima"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 13
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Projection of data \n",
      "We are only going to analyze loci that have data for at least 3 individuals in each population, and for at least one outgroup individual. Because we have more samples from Florida we use a minimum of 5 individuals (10 'chromosomes') for this population. The projection is thus [10,6,6], and we use this below to filter the loci down to a list `Floci` that will be examined. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## minimum samples required to use the locus\n",
      "proj = [10,6,6]\n",
      "\n",
      "## only examine loci w/ at least 1 outgroup sample\n",
      "## and at least two inds from each focal populations\n",
      "Floci = []   ## filtered locus list\n",
      "for loc in loci:\n",
      "    names = [i.strip().split(\" \")[0] for i in loc.strip().split(\"\\n\")[:-1]]  \n",
      "    if len(set([\">\"+i for i in outgroup]).intersection(names)) >= 1:\n",
      "        if len(set([\">\"+i for i in florida]).intersection(names)) >= proj[0]/2:\n",
      "            if len(set([\">\"+i for i in centralA]).intersection(names)) >= proj[1]/2:\n",
      "                if len(set([\">\"+i for i in cuba]).intersection(names)) >= proj[2]/2:\n",
      "                    Floci.append(loc)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Create dadi formatted file\n",
      "Now we sample from the list of `Floci` to select a single SNP from each variable locus and write this to a file.  "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## outfile location            \n",
      "outfile = open(\"oaks.dadi.snps\", 'w')\n",
      "    \n",
      "## print header\n",
      "print >>outfile, \"\\t\".join([\"REF\",\"OUT\",\n",
      "                            \"Allele1\",\"fl\",\"ca\",\"cu\",\n",
      "                            \"Allele2\",\"fl\",\"ca\",\"cu\",\n",
      "                            \"locus\",\"site\"])\n",
      "## recording values\n",
      "uncalled = 0\n",
      "calledSNP = 0\n",
      "uncalledSNP = 0\n",
      "\n",
      "## iterate over loci\n",
      "for locN, loc in enumerate(Floci):\n",
      "    ## to filter a single SNP per locus\n",
      "    singleSNP = 0\n",
      "\n",
      "    ## separate names, loci, and SNP identifier line\n",
      "    names = [i.strip().split(\" \")[0] for i in loc.strip().split(\"\\n\")[:-1]]  \n",
      "    dat = [i.strip().split(\" \")[-1] for i in loc.strip().split(\"\\n\")[:-1]]\n",
      "    snps = loc.strip().split(\"\\n\")[-1][14:]\n",
      "    \n",
      "    ## select ingroups v. outgroups\n",
      "    ihits = [names.index(\">\"+i) for i in cuba+centralA+florida \\\n",
      "             if \">\"+i in names]\n",
      "    ohits = [names.index(\">\"+i) for i in outgroup \\\n",
      "             if \">\"+i in names]\n",
      "    \n",
      "    ## select a single variable SNP from each locus\n",
      "    ## but also record how many are skipped for calculating\n",
      "    ## effective sequence length later\n",
      "    for site,char in enumerate(snps):\n",
      "        if char in [\"-\",\"*\"]:\n",
      "            ## get common (reference) alleles\n",
      "            i2 = [\"-\"+dat[k][site]+\"-\" for k in ihits if dat[k][site] not in [\"N\",\"-\"]]\n",
      "            o2 = [\"-\"+dat[k][site]+\"-\" for k in ohits if dat[k][site] not in [\"N\",\"-\"]]\n",
      "            ## filter out uninformative sites\n",
      "            b1 = any([i[1] not in ['N','-'] for i in i2])\n",
      "            b2 = any([i[1] not in ['N','-'] for i in o2])\n",
      "            if not (b1 and b2):\n",
      "                uncalled += 1\n",
      "            else:\n",
      "                ## if site is variable in the ingroups\n",
      "                if len(set(i2)) > 1:\n",
      "                    ## get the segregating alleles\n",
      "                    alleles = list(itt.chain(*[unstruct(i[1]) for i in i2]))\n",
      "                    allele1 = most_common(alleles)\n",
      "                    allele2 = most_common([i for i in alleles if i!=allele1])\n",
      "                    outg = most_common(list(itt.chain(*[unstruct(i[1]) for i in o2])))\n",
      "                    ## florida\n",
      "                    fl = [names.index(\">\"+z) for z in florida if \">\"+z in names]\n",
      "                    fldat = list(itt.chain(*[unstruct(dat[z][site]) for z in fl]))\n",
      "                    fl1,fl2 = fldat.count(allele1), fldat.count(allele2)\n",
      "                    ## central America\n",
      "                    ca = [names.index(\">\"+z) for z in centralA if \">\"+z in names]\n",
      "                    cadat = list(itt.chain(*[unstruct(dat[z][site]) for z in ca]))\n",
      "                    ca1,ca2 = cadat.count(allele1), cadat.count(allele2)\n",
      "                    ## cuba\n",
      "                    cu = [names.index(\">\"+z) for z in cuba if \">\"+z in names]\n",
      "                    cudat = list(itt.chain(*[unstruct(dat[z][site]) for z in cu]))\n",
      "                    cu1,cu2 = cudat.count(allele1), cudat.count(allele2)\n",
      "                    \n",
      "                    if not singleSNP:\n",
      "                        calledSNP += 1\n",
      "                        print >>outfile, \"\\t\".join(map(str,[\"-\"+allele1+\"-\",\n",
      "                                                        \"-\"+outg+\"-\",\n",
      "                                                        allele1, fl1, ca1, cu1,\n",
      "                                                        allele2, fl2, ca2, cu2,\n",
      "                                                        str(locN), str(site)])) \n",
      "                        singleSNP = 1\n",
      "                    else: \n",
      "                        uncalledSNP += 1\n",
      "\n",
      "outfile.close()\n",
      "\n",
      "Nloci = len(Floci)\n",
      "print \"loci\\t\",Nloci\n",
      "print \"bp\\t\",Nloci*90\n",
      "L = Nloci*90-uncalled\n",
      "print \"called bp\\t\",L\n",
      "print 'called SNP\\t', calledSNP\n",
      "print 'uncalledSNP\\t', uncalledSNP\n",
      "propcalled = calledSNP/float(calledSNP+uncalledSNP)\n",
      "print 'prop\\t', propcalled"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "loci\t7794\n",
        "bp\t701460\n",
        "called bp\t701161\n",
        "called SNP\t2043\n",
        "uncalledSNP\t836\n",
        "prop\t0.709621396318\n"
       ]
      }
     ],
     "prompt_number": 16
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### The input file looks like this"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "! head -20 oaks.dadi.snps"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "REF\tOUT\tAllele1\tfl\tca\tcu\tAllele2\tfl\tca\tcu\tlocus\tsite\r\n",
        "-C-\t-C-\tC\t19\t8\t6\tA\t1\t0\t0\t3\t56\r\n",
        "-G-\t-G-\tG\t15\t10\t6\tA\t1\t0\t0\t10\t37\r\n",
        "-G-\t-G-\tG\t20\t6\t6\tC\t2\t0\t0\t23\t11\r\n",
        "-C-\t-A-\tC\t19\t10\t5\tT\t1\t0\t1\t25\t25\r\n",
        "-T-\t-T-\tT\t13\t8\t6\tC\t1\t0\t0\t28\t78\r\n",
        "-C-\t-C-\tC\t22\t7\t6\tT\t0\t1\t0\t32\t78\r\n",
        "-T-\t-T-\tT\t16\t7\t6\tC\t2\t1\t0\t38\t65\r\n",
        "-T-\t-T-\tT\t18\t7\t6\tC\t0\t1\t0\t40\t80\r\n",
        "-G-\t-G-\tG\t9\t8\t6\tC\t13\t0\t0\t42\t29\r\n",
        "-T-\t-T-\tT\t11\t6\t6\tG\t3\t0\t0\t43\t84\r\n",
        "-T-\t-T-\tT\t10\t8\t4\tC\t2\t0\t0\t51\t85\r\n",
        "-C-\t-C-\tC\t22\t9\t6\tA\t0\t1\t0\t53\t10\r\n",
        "-T-\t-T-\tT\t12\t9\t6\tG\t0\t1\t0\t54\t28\r\n",
        "-T-\t-T-\tT\t16\t8\t6\tC\t2\t0\t0\t56\t86\r\n",
        "-T-\t-T-\tT\t15\t8\t6\tC\t1\t0\t0\t65\t40\r\n",
        "-C-\t-C-\tC\t8\t8\t6\tT\t2\t0\t0\t75\t68\r\n",
        "-C-\t-C-\tC\t11\t5\t6\tG\t1\t1\t0\t80\t6\r\n",
        "-T-\t-T-\tT\t22\t10\t4\tC\t0\t0\t2\t81\t87\r\n",
        "-C-\t-C-\tC\t14\t10\t6\tG\t4\t0\t0\t85\t10\r\n"
       ]
      }
     ],
     "prompt_number": 17
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Convert dadi snps file to SFS \n",
      "After trying several projections, and with or without corners masked, I settled on the following data set. The projection maximizes data available for the three Cuban samples, and unmasking corners incorporates a significant amount of information about alleles that are fixed between populations, since the populations used here are fairly divergent. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## parse the snps file\n",
      "dd = dadi.Misc.make_data_dict('oaks.dadi.snps')\n",
      "\n",
      "## extract the fs from this dict with data not projected \n",
      "## down to fewer than the total samples\n",
      "fs = dadi.Spectrum.from_data_dict(dd,\n",
      "                                  pop_ids=[\"fl\",\"ca\",\"cu\"],\n",
      "                                  projections=proj, \n",
      "                                  polarized=True)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 18
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## how many segregating sites after projection?\n",
      "print fs.S()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1626.19720629\n"
       ]
      }
     ],
     "prompt_number": 19
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "# Define three _IM_-type models  \n",
      "Models `IM1` and `IM2` are isolation-migration models for a 3 population tree in which Cuba is derived from one lineage or the other. The `admix` model has the Cuban population formed as an instantaneous fusion of the other two populations, after which it does not exchange  gene flow with the parent populations. A separate population size is estimated for each population. Migration is set to zero between Florida and Central America. "
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "![](https://raw.githubusercontent.com/dereneaton/virentes/master/models_simple.png)"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Model 1: Island conversion model"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## scenario ((florida,cuba),centralA);\n",
      "\n",
      "def IM_split1(params, ns, pts):\n",
      "    ## parse params\n",
      "    N1,N2,N3,T2,T1,m13,m31,m23,m32 = params\n",
      "    \n",
      "    ## create a search grid\n",
      "    xx = dadi.Numerics.default_grid(pts)\n",
      "    \n",
      "    ## create ancestral pop\n",
      "    phi = dadi.PhiManip.phi_1D(xx)\n",
      "    \n",
      "    ## split ancestral pop into two species\n",
      "    phi = dadi.PhiManip.phi_1D_to_2D(xx,phi)\n",
      "    \n",
      "    ## allow drift to occur along each of these branches\n",
      "    phi = dadi.Integration.two_pops(phi, xx, T2, nu1=N1, nu2=N2, m12=0., m21=0.)\n",
      "    \n",
      "    ## split pop1 into pops 1 and 3\n",
      "    phi = dadi.PhiManip.phi_2D_to_3D_split_1(xx, phi)\n",
      "\n",
      "    ## allow drift and migration to occur along these branches\n",
      "    phi = dadi.Integration.three_pops(phi, xx, T1,\n",
      "                                       nu1=N1, nu2=N2, nu3=N3,  \n",
      "                                       m12=0.0, m13=m13, \n",
      "                                       m21=0.0, m23=m23,\n",
      "                                       m31=m31, m32=m32)\n",
      "    ## simulate the fs\n",
      "    fs = dadi.Spectrum.from_phi(phi,ns,(xx,xx,xx), \n",
      "                                pop_ids=['fl', 'ca', 'cu'])\n",
      "    return fs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 8
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Model 2: Island conversion model"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## scenario ((centralA,cuba),florida);\n",
      "\n",
      "def IM_split2(params, ns, pts):\n",
      "    ## parse params\n",
      "    N1,N2,N3,T2,T1,m13,m31,m23,m32 = params\n",
      "    \n",
      "    ## create a search grid\n",
      "    xx = dadi.Numerics.default_grid(pts)\n",
      "    \n",
      "    ## create ancestral pop\n",
      "    phi = dadi.PhiManip.phi_1D(xx)\n",
      "    \n",
      "    ## split ancestral pop into two species\n",
      "    phi = dadi.PhiManip.phi_1D_to_2D(xx,phi)\n",
      "    \n",
      "    ## allow drift to occur along each of these branches\n",
      "    phi = dadi.Integration.two_pops(phi, xx, T2, nu1=N1, nu2=N2, m12=0., m21=0.)\n",
      "    \n",
      "    ## split pop2 into pops 2 and 3\n",
      "    phi = dadi.PhiManip.phi_2D_to_3D_split_2(xx, phi)\n",
      "\n",
      "    ## allow drift and migration to occur along these branches\n",
      "    phi = dadi.Integration.three_pops(phi, xx, T1,\n",
      "                                      nu1=N1, nu2=N2, nu3=N3,  \n",
      "                                      m12=0.0, m13=m13, \n",
      "                                      m21=0.0, m23=m23,\n",
      "                                      m31=m31, m32=m32)\n",
      "    ## simulate the fs\n",
      "    fs = dadi.Spectrum.from_phi(phi,ns,(xx,xx,xx), \n",
      "                                pop_ids=['fl', 'ca', 'cu'])\n",
      "    return fs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 9
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Model 3: instantaneous admixture"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## scenario hybrid speciation forms cuba.\n",
      "\n",
      "def admix(params, ns, pts):\n",
      "    ## parse params\n",
      "    N1,N2,N3,T2,T1,f = params\n",
      "    \n",
      "    ## create a search grid\n",
      "    xx = dadi.Numerics.default_grid(pts)\n",
      "    \n",
      "    ## make ancestral pop that splits into two\n",
      "    phi = dadi.PhiManip.phi_1D(xx)\n",
      "    phi = dadi.PhiManip.phi_1D_to_2D(xx,phi)\n",
      "    \n",
      "    ## allow drift to occur along each of these branches\n",
      "    phi = dadi.Integration.two_pops(phi, xx, T2, nu1=N1, nu2=N2, m12=0., m21=0.)\n",
      "    \n",
      "    ## create pop 3 from a mixture of 1 and 2\n",
      "    phi = dadi.PhiManip.phi_2D_to_3D_admix(phi, f, xx, xx, xx)\n",
      "\n",
      "    ## allow drift and migration to occur along these branches\n",
      "    ## cuba population shrinks in size after divergence\n",
      "    phi = dadi.Integration.three_pops(phi, xx, T1,\n",
      "                                      nu1=N1, nu2=N2, nu3=N3,  \n",
      "                                      m12=0.0, m13=0.0, \n",
      "                                      m21=0.0, m23=0.0,\n",
      "                                      m31=0.0, m32=0.0)\n",
      "    ## simulate the fs\n",
      "    fs = dadi.Spectrum.from_phi(phi,ns,(xx,xx,xx),\n",
      "                                pop_ids=['fl', 'ca', 'cu'])\n",
      "    return fs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 10
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "# Model fitting\n",
      "### Starting parameters and limits  "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## sample sizes\n",
      "ns = fs.sample_sizes\n",
      "\n",
      "## points used for exrapolation\n",
      "pts_l = [12, 20, 32]\n",
      "\n",
      "## starting values for params\n",
      "N1 = N2 = N3 = 2.0\n",
      "T2 = 2.0\n",
      "T1 = 2.0\n",
      "m13 = m31 = m23 = m32 = 0.1\n",
      "f  = 0.4999\n",
      "\n",
      "## create starting parameter sets\n",
      "params_IM    = np.array([N1,N2,N3,T2,T1,m13,m31,m23,m32])\n",
      "params_admix = np.array([N1,N2,N3,T2,T1,f])\n",
      "\n",
      "## search limits \n",
      "upper_IM    = [10.0,10.0,10.0, 10.0,10.0, 25.0, 25.0, 25.0, 25.0]\n",
      "lower_IM    = [1e-3,1e-3,1e-3, 1e-3,1e-3, 1e-5, 1e-5, 1e-5, 1e-5]\n",
      "\n",
      "upper_admix = [10.0,10.0,10.0, 10.0,10.0, 0.9999]\n",
      "lower_admix = [1e-3,1e-3,1e-3, 1e-3,1e-3, 0.0001]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 20
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Python script for model fitting \n",
      "\n",
      "I wrote a separate Python script for each model in which the parameters were estimated using the commands below. This was done 10 times for each model from different starting positions. These were executed in separate Python shells because optimization sometimes freezes in bad parameter space, and because it allowed easy parallelization. These scripts are available in the `git` repository as `dadi_m1X.py`, `dadi_m2X.py`, and `dadi_m3X.py`. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%bash\n",
      "## example of running code\n",
      "## python analysis_dadi/dadi_m1X.py"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#### Example code for optimization in script dadi_m1X.py\n",
      "## model = IM_split1\n",
      "## params= params_IM\n",
      "## uppers= upper_IM\n",
      "## lowers= lower_IM\n",
      "## maxiters=10\n",
      "\n",
      "#### Create extrapolating function \n",
      "## Func = dadi.Numerics.make_extrap_log_func(model)\n",
      "\n",
      "#### Perturb start values\n",
      "## p0 = dadi.Misc.perturb_params(params, fold=2., \n",
      "##                              upper_bound=uppers,\n",
      "##                              lower_bound=lowers)\n",
      "\n",
      "#### optimize from starting position\n",
      "## args = [p0, fs, Func, pts_l, lowers, uppers, 0, maxiters]\n",
      "## popt = dadi.Inference.optimize_log_lbfgsb(*args)\n",
      "\n",
      "#### return the likelihood of the data under these parameters \n",
      "## mod = Func(popt, ns, pts_l)\n",
      "## ll_opt = dadi.Inference.ll_multinom(mod, fs)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Optimized best fit from 10 random starting positions"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## optimized at projection (10,6,6); extrap (12,20,32); S=1626\n",
      "ml1 = [2.5013156716397473, 0.68871161421841698, 0.076807020757753017,\n",
      "       0.70851628972828207, 0.14901468477012897, 0.14349173980106683,\n",
      "       12.58678666702065, 1.2625236526206685, 21.036230092460865, -543.08153591196356]\n",
      "\n",
      "ml2 = [2.412336866389539, 0.66060145092919675, 0.23061219403870617,\n",
      "       0.71073109877391338, 0.084218967415615298, 0.138041546493126,\n",
      "       6.036453130870238, 4.09467065466802, 0.0031179372963987696, -541.90248015938994]\n",
      "\n",
      "ml3 = [2.2562645699939852, 0.71705301735090954, 0.14307134998023263,\n",
      "       0.57976328038133673, 0.025790751880146609, 0.38345451331222341,\n",
      "       -555.34376821105138]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "indexnames = [\"N1\",\"N2\",\"N3\",\"T2\",\"T1\",\"m13\",\"m31\",\"m23\",\"m32\",\"LL\"]\n",
      "ML1 = pd.Series(ml1, index = indexnames)\n",
      "ML2 = pd.Series(ml2, index = indexnames)\n",
      "ML3 = pd.Series(ml3, index=[\"N1\",\"N2\",\"N3\",\"T2\",\"T1\",\"f\",\"LL\"])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = pd.DataFrame([ML1,ML2,ML3],index=[\"ML1\",\"ML2\",\"ML3\"]).T\n",
      "print data"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "         ML1      ML2      ML3\n",
        "LL  -543.082 -541.902 -555.344\n",
        "N1     2.501    2.412    2.256\n",
        "N2     0.689    0.661    0.717\n",
        "N3     0.077    0.231    0.143\n",
        "T1     0.149    0.084    0.026\n",
        "T2     0.709    0.711    0.580\n",
        "f        NaN      NaN    0.383\n",
        "m13    0.143    0.138      NaN\n",
        "m23    1.263    4.095      NaN\n",
        "m31   12.587    6.036      NaN\n",
        "m32   21.036    0.003      NaN\n",
        "\n",
        "[11 rows x 3 columns]\n"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "----------------   \n",
      "\n",
      "## Converting units\n",
      "\n",
      "### Get  $\\theta$ values"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "## optimized grid\n",
      "pts_l = [12,20,32]\n",
      "\n",
      "# The optimal value of theta given the model.\n",
      "Func = dadi.Numerics.make_extrap_log_func(IM_split1)\n",
      "model1_AFS = Func(ML1.tolist()[:-1],ns,pts_l)\n",
      "theta1 = dadi.Inference.optimal_sfs_scaling(model1_AFS, fs)\n",
      "\n",
      "# The optimal value of theta given the model.\n",
      "Func = dadi.Numerics.make_extrap_log_func(IM_split2)\n",
      "model2_AFS = Func(ML2.tolist()[:-1],ns,pts_l)\n",
      "theta2 = dadi.Inference.optimal_sfs_scaling(model2_AFS, fs)\n",
      "\n",
      "# The optimal value of theta given the model.\n",
      "Func = dadi.Numerics.make_extrap_log_func(admix)\n",
      "model3_AFS = Func(ML3.tolist()[:-1],ns,pts_l)\n",
      "theta3 = dadi.Inference.optimal_sfs_scaling(model3_AFS, fs)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 21
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### ML estimates in converted units"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "###### Fixed params\n",
      "mu = 2.5e-9         ## from Populus article by xxyy\n",
      "gentime = 30.0   ## (years) from Gugger & Cavender-Bares\n",
      "\n",
      "models = [ML1,ML2,ML3]\n",
      "thetas = [theta1,theta2,theta3]\n",
      "for model,theta in zip(models,thetas):\n",
      "    print \"\\n-----model %s-------\" % str(thetas.index(theta)+1)\n",
      "    ## Population sizes\n",
      "    Nref = theta/(4*mu*L)\n",
      "    print \"theta =\\t\",theta, \"(\"+str(theta/Nloci)+\")\"\n",
      "    print \"Nref =\\t\",round(Nref)\n",
      "    print \"N1 =\\t\",round(Nref*model[\"N1\"])\n",
      "    print \"N2 =\\t\",round(Nref*model[\"N2\"])\n",
      "    print \"N3 =\\t\",round(Nref*model[\"N3\"])\n",
      "    ## Divergence times\n",
      "    T1  = model[\"T1\"]\n",
      "    T12 = model[\"T1\"] + model[\"T2\"]\n",
      "    print 'T1  =\\t',round( ( T1*2*Nref * gentime)/1e6, 3), 'Mya'\n",
      "    print 'T12 =\\t',round( (T12*2*Nref * gentime)/1e6, 3), 'Mya'\n",
      "    ## Migration\n",
      "    if 'f' in model.index:\n",
      "        print 'f   =\\t', model['f']\n",
      "        print '1-f =\\t', 1-(model[\"f\"])\n",
      "    else:\n",
      "        print 'm13 =\\t', (model[\"m13\"])/(2*Nref) * 1e3\n",
      "        print 'm31 =\\t', (model[\"m31\"])/(2*Nref) * 1e3\n",
      "        print 'm23 =\\t', (model[\"m23\"])/(2*Nref) * 1e3\n",
      "        print 'm32 =\\t', (model[\"m32\"])/(2*Nref) * 1e3"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "-----model 1-------\n",
        "theta =\t249.587274958 (0.0320230016626)\n",
        "Nref =\t35596.0\n",
        "N1 =\t89038.0\n",
        "N2 =\t24516.0\n",
        "N3 =\t2734.0\n",
        "T1  =\t0.318 Mya\n",
        "T12 =\t1.831 Mya\n",
        "m13 =\t0.00201554369684\n",
        "m31 =\t0.176799156281\n",
        "m23 =\t0.0177339238738\n",
        "m32 =\t0.295483496311\n",
        "\n",
        "-----model 2-------\n",
        "theta =\t256.759934829 (0.0329432813484)\n",
        "Nref =\t36619.0\n",
        "N1 =\t88338.0\n",
        "N2 =\t24191.0\n",
        "N3 =\t8445.0\n",
        "T1  =\t0.185 Mya\n",
        "T12 =\t1.747 Mya\n",
        "m13 =\t0.00188482188323\n",
        "m31 =\t0.0824218450692\n",
        "m23 =\t0.0559087104615\n",
        "m32 =\t4.25723747386e-05\n",
        "\n",
        "-----model 3-------\n",
        "theta =\t282.456200226 (0.0362402104472)\n",
        "Nref =\t40284.0\n",
        "N1 =\t90892.0\n",
        "N2 =\t28886.0\n",
        "N3 =\t5763.0\n",
        "T1  =\t0.062 Mya\n",
        "T12 =\t1.464 Mya\n",
        "f   =\t0.383454513312\n",
        "1-f =\t0.616545486688\n"
       ]
      }
     ],
     "prompt_number": 35
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "-------------   \n",
      "\n",
      "## Residual plots  \n",
      "\n",
      "### Model 1"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.rcParams['figure.figsize'] = (6,6)\n",
      "dadi.Plotting.plot_3d_comp_multinom(model1_AFS, fs, vmin=0.1, resid_range=3,\n",
      "                                    pop_ids =('fl', 'ca', 'cu'))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAbYAAAGbCAYAAACoO7WYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIyIiU/l/Zi9SVfdObTcAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0xd857a10>"
       ]
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Model 2"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.rcParams['figure.figsize'] = (6,6)\n",
      "dadi.Plotting.plot_3d_comp_multinom(model2_AFS, fs, vmin=0.1, resid_range=3,\n",
      "                                    pop_ids =('fl', 'ca', 'cu'))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAbYAAAGbCAYAAACoO7WYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFBY2IiIylP8DpWvMvv0fUhUAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0xeeaf590>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0x1087eb90>"
       ]
      }
     ],
     "prompt_number": 23
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Model 3"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.rcParams['figure.figsize'] = (6,6)\n",
      "dadi.Plotting.plot_3d_comp_multinom(model3_AFS, fs, vmin=0.1, resid_range=3,\n",
      "                                    pop_ids =('fl', 'ca','cu'))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAbYAAAGbCAYAAACoO7WYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10897d90>"
       ]
      }
     ],
     "prompt_number": 24
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "---------------   \n",
      "\n",
      "## Parametric bootstrapping from simulated data in `ms`. \n",
      "\n",
      "We generate simulated data sets under the same demographic scenarios as in the dadi models. We save each of the simulated data sets because these will also be used for model selection below. "
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## Functions for simulating data in `ms`"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#############################################################\n",
      "### IM_split 1 \n",
      "def simSFS_model1(theta, N1,N2,N3,mig,T1, T12):\n",
      "    R1,R2,R3 = np.random.random_integers(0,9999999,3)\n",
      "    command = \"\"\" ms 22 7794 \\\n",
      "                     -t %(theta)f \\\n",
      "                     -I 3 10 6 6 \\\n",
      "                     -n 1 %(N1)f \\\n",
      "                     -n 2 %(N2)f \\\n",
      "                     -n 3 %(N3)f \\\n",
      "                     -en %(T1)f 1 %(N1)f \\\n",
      "                     -ma %(mig)s \\\n",
      "                     -ej %(T1)f 3 1 \\\n",
      "                     -ej %(T12)f 2 1 \\\n",
      "                     -en %(T12)f 1 1 \\\n",
      "                     -seeds %(R1)f %(R2)f %(R3)f \"\"\"\n",
      "    sub_dict = {'theta':theta, 'N1':N1, 'N2':N2, 'N3':N3,\n",
      "                'T1':T1, 'mig':\" \".join(mig),\n",
      "                'R1':R1, 'R2':R2, 'R3':R3}\n",
      "    mscommand = command % sub_dict\n",
      "    fs = dadi.Spectrum.from_ms_file(os.popen(mscommand), average=False)\n",
      "    return fs\n",
      "\n",
      "###############################################################\n",
      "## IM_split 2\n",
      "def simSFS_model2(theta, N1,N2,N3,mig,T1, T12):\n",
      "    R1,R2,R3 = np.random.random_integers(0,9999999,3)\n",
      "    command = \"\"\" ms 22 7794 \\\n",
      "                     -t %(theta)f \\\n",
      "                     -I 3 10 6 6 \\\n",
      "                     -n 1 %(N1)f \\\n",
      "                     -n 2 %(N2)f \\\n",
      "                     -n 3 %(N3)f \\\n",
      "                     -en %(T1)f 2 %(N2)f \\\n",
      "                     -ma %(mig)s \\\n",
      "                     -ej %(T1)f 3 2 \\\n",
      "                     -ej %(T12)f 2 1 \\\n",
      "                     -en %(T12)f 1 1 \\\n",
      "                     -seeds %(R1)f %(R2)f %(R3)f \"\"\"\n",
      "    sub_dict = {'theta':theta, 'N1':N1, 'N2':N2, 'N3':N3,\n",
      "                'T1':T1, 'mig':\" \".join(mig),\n",
      "                'R1':R1, 'R2':R2, 'R3':R3}\n",
      "    mscommand = command % sub_dict\n",
      "    fs = dadi.Spectrum.from_ms_file(os.popen(mscommand), average=False)\n",
      "    return fs\n",
      "\n",
      "#################################################################\n",
      "## Admix\n",
      "def simSFS_model3(theta, N1,N2,N3,T1, T12,f):\n",
      "    R1,R2,R3 = np.random.random_integers(0,9999999,3)\n",
      "    command = \"\"\" ms 22 7794 \\\n",
      "                     -t %(theta)f \\\n",
      "                     -I 3 10 6 6 \\\n",
      "                     -n 1 %(N1)f \\\n",
      "                     -n 2 %(N2)f \\\n",
      "                     -n 3 %(N3)f \\\n",
      "                     -en %(T1)f 2 %(N2)f \\\n",
      "                     -en %(T1)f 1 %(N1)f \\\n",
      "                     -es %(T1)f 3 %(f)f \\\n",
      "                     -ej %(T1)f 4 2 \\\n",
      "                     -ej %(T1)f 3 1 \\\n",
      "                     -ej %(T12)f 2 1 \\\n",
      "                     -en %(T12)f 1 1 \\\n",
      "                     -seeds %(R1)f %(R2)f %(R3)f \"\"\"\n",
      "    sub_dict = {'theta':theta, 'N1':N1, 'N2':N2, 'N3':N3,\n",
      "                'T1':T1, 'f':f,\n",
      "                'R1':R1, 'R2':R2, 'R3':R3}\n",
      "    mscommand = command % sub_dict\n",
      "    fs = dadi.Spectrum.from_ms_file(os.popen(mscommand), average=False)\n",
      "    return fs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 636
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Estimated parameters"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Execute Python scripts with the above code in a separate shell\n",
      "This creates 200 parametric bootstrap data sets for each of the three models, and re-estimates parameters under each of the three models. The sfs for each bootstrap data set is saved in a directory called pboots/. The Python scripts to execute this are in the `git` repository as `dadi_m1boot_fit3.py`, `dadi_m2boot_fit3.py`, and `dadi_m3boot_fit3.py`. Nine output files were created and in are in the `git` repo. For example, the optimized parameters for a data set simulated under model 1 and fit using model 2 is saved to the file `dadi_m1_f2_boots.txt`."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%bash\n",
      "## commented out b/c it takes a long time to run\n",
      "\n",
      "## for i in $(seq 1 200);\n",
      "##   do python analysis_dadi/dadi_m1boot_fit3.py\n",
      "## done\n",
      "\n",
      "## for i in $(seq 1 200);\n",
      "##   do python analysis_dadi/dadi_m2boot_fit3.py\n",
      "## done\n",
      "\n",
      "## for i in $(seq 1 100);\n",
      "##   do python analysis_dadi/dadi_m3boot_fit3.py\n",
      "## done"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 510
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Calculate biased-corrected confidence intervals"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "###### Fixed params\n",
      "mu = 2.5e-9         ## from Populus genome article \n",
      "gentime = 30.0     ## (years) from Gugger & Cavender-Bares\n",
      "\n",
      "def makedict(bootfile, theta):    \n",
      "    ## parse file\n",
      "    DD = {}\n",
      "    infile = open(bootfile).readlines()\n",
      "    lines = [i.strip().split(\"\\t\") for i in infile]\n",
      "    bootorder = [int(i[0].replace(\"boot\",\"\")) for i in lines]\n",
      "    elements = [\"N1\",\"N2\",\"N3\",\"T2\",\"T1\"]\n",
      "    for el,li in zip(elements, range(1,6)):\n",
      "        DD[el] = [float(i[li]) for i in lines]\n",
      "    if len(lines[0]) > 8:\n",
      "        elements = [\"m13\",\"m31\",\"m23\",\"m32\",\"LL\"]\n",
      "        for el,li in zip(elements, range(6,11)):\n",
      "            DD[el] = [float(i[li]) for i in lines]\n",
      "    else:\n",
      "        DD['f'] = [float(i[6]) for i in lines]\n",
      "        DD['LL'] = [float(i[7]) for i in lines]\n",
      "    DD['sLL'] = [DD[\"LL\"][i-1] for i in bootorder[:60]]\n",
      "    return DD"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 36
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Model 1"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"analysis_dadi/dadi_m1_f1_boots.txt\", theta1)\n",
      "\n",
      "#############################\n",
      "Nref = theta1/(4*mu*L)\n",
      "#############################\n",
      "\n",
      "print \"Model 1\"\n",
      "for param in [\"N1\",\"N2\",\"N3\"]:\n",
      "    m = np.mean([(i*Nref) for i in DD[param]])\n",
      "    s = np.std([(i*Nref) for i in DD[param]])\n",
      "    o = (ML1[param]*Nref)\n",
      "    unbiased = o-(m-o)\n",
      "    print param, max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "\n",
      "#############################\n",
      "m = np.mean([(i*Nref*2*gentime)/1e6 for i in DD[\"T1\"]])\n",
      "s = np.std([(i*Nref*2*gentime)/1e6 for i in DD[\"T1\"]])\n",
      "o = (ML1[\"T1\"]*Nref*2*gentime)/1e6\n",
      "unbiased = o-(m-o)\n",
      "print \"T1\", max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "m = np.mean([(i*Nref*2*gentime)/1e6 for i in (DD[\"T1\"]+DD[\"T2\"])])\n",
      "s = np.std([(i*Nref*2*gentime)/1e6 for i in (DD[\"T1\"]+DD[\"T2\"])])\n",
      "o = ((ML1[\"T1\"]+ML1[\"T2\"])*Nref*2*gentime)/1e6\n",
      "unbiased = o-(m-o)\n",
      "print \"T12\", max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "\n",
      "    \n",
      "#############################\n",
      "for param in [\"m13\",\"m31\",\"m23\",\"m32\"]:\n",
      "    m = np.mean([i/(2*Nref) for i in DD[param]])\n",
      "    s = np.std([i/(2*Nref) for i in DD[param]])\n",
      "    o = (ML1[param]/(2*Nref))\n",
      "    unbiased = o-(m-o)\n",
      "    print param, max(0,(unbiased-1.96*s)*1e3), o*1e3, (unbiased+1.96*s)*1e3\n",
      "    \n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML1[\"LL\"]\n",
      "unbiased = o-(m-o)\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Model 1\n",
        "N1 71475.9003173 89037.5480517 100269.854037\n",
        "N2 18585.0359322 24515.5755989 29469.4608854\n",
        "N3 0 2734.04467855 5301.53276347\n",
        "T1 0 0.318262160336 0.896697430149\n",
        "T12 1.43013516666 1.8314950699 3.99576741889\n",
        "m13 0 0.00201554369684 0.00779580837604\n",
        "m31 0.0204317889296 0.176799156281 0.338742961336\n",
        "m23 0.00280410835358 0.0177339238738 0.0298659118644\n",
        "m32 0.0155455991765 0.295483496311 0.519561371751\n",
        "LL -611.90667302 -543.081535912 -547.307616243\n"
       ]
      }
     ],
     "prompt_number": 37
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## Model 2"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"analysis_dadi/dadi_m2_f2_boots.txt\", theta2)\n",
      "\n",
      "#############################\n",
      "Nref = theta2/(4*mu*L)\n",
      "#############################\n",
      "print \"Model 2\"\n",
      "for param in [\"N1\",\"N2\",\"N3\"]:\n",
      "    m = np.mean([(i*Nref) for i in DD[param]])\n",
      "    s = np.std([(i*Nref) for i in DD[param]])\n",
      "    o = (ML2[param]*Nref)\n",
      "    unbiased = o-(m-o)\n",
      "    print param, max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "\n",
      "#############################\n",
      "m = np.mean([(i*Nref*2*gentime)/1e6 for i in DD[\"T1\"]])\n",
      "s = np.std([(i*Nref*2*gentime)/1e6 for i in DD[\"T1\"]])\n",
      "o = (ML2[\"T1\"]*Nref*2*gentime)/1e6\n",
      "unbiased = o-(m-o)\n",
      "print \"T1\", max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "m = np.mean([(i*Nref*2*gentime)/1e6 for i in (DD[\"T1\"]+DD[\"T2\"])])\n",
      "s = np.std([(i*Nref*2*gentime)/1e6 for i in (DD[\"T1\"]+DD[\"T2\"])])\n",
      "o = ((ML2[\"T1\"]+ML2[\"T2\"])*Nref*2*gentime)/1e6\n",
      "unbiased = o-(m-o)\n",
      "print \"T12\", max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "\n",
      "    \n",
      "#############################\n",
      "for param in [\"m13\",\"m31\",\"m23\",\"m32\"]:\n",
      "    m = np.mean([i/(2*Nref) for i in DD[param]])\n",
      "    s = np.std([i/(2*Nref) for i in DD[param]])\n",
      "    o = (ML2[param]/(2*Nref))\n",
      "    unbiased = o-(m-o)\n",
      "    print param, max(0,(unbiased-1.96*s)*1e3), o*1e3, (unbiased+1.96*s)*1e3\n",
      "    \n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML2[\"LL\"]\n",
      "unbiased = o-(m-o)\n",
      "#print \"LL\", unbiased-1.96*s, o, unbiased+1.96*s\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Model 2\n",
        "N1 69722.5763103 88337.9789521 100024.420796\n",
        "N2 17818.7840481 24190.7330112 29340.4289636\n",
        "N3 2379.47192323 8444.84674878 13458.644633\n",
        "T1 0.0363476312861 0.185042150819 0.305737745991\n",
        "T12 1.18583067533 1.74662875306 3.99630587332\n",
        "m13 0 0.00188482188323 0.00472397460442\n",
        "m31 0.00752720476581 0.0824218450692 0.148066108791\n",
        "m23 0.0242008829239 0.0559087104615 0.0867269369484\n",
        "m32 0 4.25723747386e-05 0.00426563585615\n",
        "LL -605.553218248 -541.902480159 -542.596028013\n"
       ]
      }
     ],
     "prompt_number": 38
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## Model 3"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"analysis_dadi/dadi_m3_f3_boots.txt\", theta2)\n",
      "\n",
      "#############################\n",
      "Nref = theta3/(4*mu*L)\n",
      "#############################\n",
      "print \"Model 3\"\n",
      "for param in [\"N1\",\"N2\",\"N3\"]:\n",
      "    m = np.mean([(i*Nref) for i in DD[param]])\n",
      "    s = np.std([(i*Nref) for i in DD[param]])\n",
      "    o = (ML3[param]*Nref)\n",
      "    unbiased = o-(m-o)\n",
      "    print param, max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "\n",
      "#############################\n",
      "m = np.mean([(i*Nref*2*gentime)/1e6 for i in DD[\"T1\"]])\n",
      "s = np.std([(i*Nref*2*gentime)/1e6 for i in DD[\"T1\"]])\n",
      "o = (ML3[\"T1\"]*Nref*2*gentime)/1e6\n",
      "unbiased = o-(m-o)\n",
      "print \"T1\", max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "m = np.mean([(i*Nref*2*gentime)/1e6 for i in (DD[\"T1\"]+DD[\"T2\"])])\n",
      "s = np.std([(i*Nref*2*gentime)/1e6 for i in (DD[\"T1\"]+DD[\"T2\"])])\n",
      "o = ((ML3[\"T1\"]+ML3[\"T2\"])*Nref*2*gentime)/1e6\n",
      "unbiased = o-(m-o)\n",
      "print \"T12\", max(0,unbiased-1.96*s), o, unbiased+1.96*s\n",
      "\n",
      "    \n",
      "#############################\n",
      "m = np.mean(DD[\"f\"])\n",
      "s = np.std(DD[\"f\"])\n",
      "o = (ML3['f'])\n",
      "unbiased = o-(m-o)\n",
      "print 'f', unbiased-1.96*s, o, unbiased+1.96*s\n",
      "    \n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML3[\"LL\"]\n",
      "unbiased = o-(m-o)\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Model 3\n",
        "N1 70873.444192 90891.5237933 104798.634546\n",
        "N2 22628.9371512 28885.8151896 33641.5723239\n",
        "N3 342.65514223 5763.49652599 10702.788427\n",
        "T1 0.00234647793955 0.0623373899321 0.114684927723\n",
        "T12 0.805824744774 1.46365092454 3.53834571718\n",
        "f 0.337902679016 0.383454513312 0.417653307748\n",
        "LL -632.794946497 -555.343768211 -566.776149859\n"
       ]
      }
     ],
     "prompt_number": 39
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Report 95% range of LL"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"dadi_m1_f3_boots.txt\", theta2)\n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML1[\"LL\"]\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL -629.472792108 -543.081535912 -562.55636251\n"
       ]
      }
     ],
     "prompt_number": 506
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"dadi_m1_f2_boots.txt\", theta2)\n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML1[\"LL\"]\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL -618.44626393 -543.081535912 -550.43080575\n"
       ]
      }
     ],
     "prompt_number": 507
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"dadi_m2_f1_boots.txt\", theta2)\n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML2[\"LL\"]\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL -610.625740669 -541.902480159 -548.678268059\n"
       ]
      }
     ],
     "prompt_number": 508
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"dadi_m2_f3_boots.txt\", theta2)\n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML2[\"LL\"]\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL -626.694918671 -541.902480159 -562.578122564\n"
       ]
      }
     ],
     "prompt_number": 509
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"dadi_m3_f1_boots.txt\", theta2)\n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML3[\"LL\"]\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL -633.253767231 -555.343768211 -568.236956872\n"
       ]
      }
     ],
     "prompt_number": 510
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "DD = makedict(\"dadi_m3_f2_boots.txt\", theta2)\n",
      "m = np.mean(DD[\"LL\"])\n",
      "s = np.std(DD[\"LL\"])\n",
      "o = ML3[\"LL\"]\n",
      "print \"LL\", m-1.96*s, o, m+1.96*s"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL -638.846613236 -555.343768211 -571.185797893\n"
       ]
      }
     ],
     "prompt_number": 511
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "------------------   \n",
      "# Monte Carlo model selection:"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Compare the area of overlap of log-likelihood ratio statistics $\\delta$ \n",
      "$\\delta$ = -2\\(log L$_2$ - log L$_1$)  \n",
      "\n",
      "following Boettiger et al. (2012)"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Can we reject the simplest model (model 3)?"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Let's begin by assuming that model 3 is true (the null model). We ask how does the observed difference in log likelihoods for the fit of our empirical data to the two models ($\\delta_{32}$ and $\\delta_{31}$) compare to the difference in log likelihood fits when the data are simulated under model 3 ($\\delta_{32sim3}$, $\\delta_{31sim3}$)?"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta32obs = -2.*(ML3[\"LL\"] - ML2[\"LL\"])\n",
      "delta31obs = -2.*(ML3[\"LL\"] - ML1[\"LL\"])\n",
      "\n",
      "print 'LL model3:', ML3[\"LL\"]\n",
      "print 'LL model2:', ML2[\"LL\"]\n",
      "print 'LL model1:', ML1[\"LL\"]\n",
      "print 'delta32 observed:', delta32obs\n",
      "print 'delta31 observed:', delta31obs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL model3: -555.343768211\n",
        "LL model2: -541.902480159\n",
        "LL model1: -543.081535912\n",
        "delta32 observed: 26.8825761033\n",
        "delta31 observed: 24.5244645982\n"
       ]
      }
     ],
     "prompt_number": 609
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Read in the optimized bootstrap fits. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "W = ['title','N1','N2','N3','T2','T1','m13','m31','m23','m32','LL']\n",
      "LL31 = pd.read_table(\"dadi_m3_f1_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]\n",
      "\n",
      "W = ['title','N1','N2','N3','T2','T1','m13','m31','m23','m32','LL']\n",
      "LL32 = pd.read_table(\"dadi_m3_f2_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]\n",
      "\n",
      "W = ['title','N1','N2','N3','T2','T1','f','LL']\n",
      "LL33 = pd.read_table(\"dadi_m3_f3_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 610
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Measure the distribution of LL differences on data simulated under model 3"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta32sim3 = -2.*(LL33-LL32)\n",
      "delta31sim3 = -2.*(LL33-LL31)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 611
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Plot the observed and simulated LL differences between models"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.rcParams['figure.figsize'] = (6,4)\n",
      "\n",
      "sns.kdeplot(delta32sim3, shade=True, label=\"$\\delta_{32}$\")\n",
      "sns.kdeplot(delta31sim3, shade=True, label=\"$\\delta_{31}$\")\n",
      "sns.despine()\n",
      "\n",
      "plt.vlines(delta32obs, 0, 0.05, lw=3, color='#262626', \n",
      "           label='$\\delta_{32obs}$', linestyles='dashed')\n",
      "plt.vlines(delta31obs, 0, 0.05, lw=3, color='#F65655', \n",
      "           label='$\\delta_{31obs}$', linestyles='dashed')\n",
      "plt.title(\"model 3 sims\")\n",
      "plt.legend(loc=2)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 551,
       "text": [
        "<matplotlib.legend.Legend at 0x22dad090>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEYCAYAAACtEtpmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x22d9ae90>"
       ]
      }
     ],
     "prompt_number": 551
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Is the difference significantly outside the expectation?\n",
      "The proportion of sims greater than the observed delta values is 0. \n",
      "This is effectively a P-value for the probability that a difference at least this large would be seen under model 3. A more accurate way, though, is to calculate $\\delta^*$, the upper 95% or 99% percentile of the simulated distribution, and ask if our observed delta is greater than this. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta32star = np.percentile(delta32sim3, 99)\n",
      "delta31star = np.percentile(delta31sim3, 99)\n",
      "print 'delta32star', delta32star\n",
      "print 'delta31star', delta31star\n",
      "print 'deltaobs', delta32obs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "delta32star 3.907066664\n",
        "delta31star 9.00921333\n",
        "deltaobs 26.8825761033\n"
       ]
      }
     ],
     "prompt_number": 604
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Can we reject model 3 for either of the more complex models? \n",
      "At a false positive rate of 0.05, Yes definitely, because observed $\\delta$ >> $\\delta^*$"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print delta32obs > delta32star\n",
      "print delta31obs > delta31star"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "True\n",
        "True\n"
       ]
      }
     ],
     "prompt_number": 553
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## Power to reject model 3"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "We next need to ask what is the probability that we correctly reject model 3 when the data comes from model 2 or model 1? This is a concern because the greater complexity of these models will bias us towards usually thinking they are better. "
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Load bootstrap data simulated under model 2."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "W = ['title','N1','N2','N3','T2','T1','m13','m31','m23','m32','LL']\n",
      "LL21 = pd.read_table(\"dadi_m2_f1_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]\n",
      "\n",
      "W = ['title','N1','N2','N3','T2','T1','m13','m31','m23','m32','LL']\n",
      "LL22 = pd.read_table(\"dadi_m2_f2_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]\n",
      "\n",
      "W = ['title','N1','N2','N3','T2','T1','f','LL']\n",
      "LL23 = pd.read_table(\"dadi_m2_f3_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 554
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Calculate $\\delta_{32}$ under model 2 sims"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta32sim2 = -2.*(LL23-LL22)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 615
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Compare $\\delta_{32}$ simulated under both models"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.rcParams['figure.figsize'] = (6,4)\n",
      "plt.gcf().subplots_adjust(bottom=0.20, left=0.15)\n",
      "\n",
      "sns.set_style(\"ticks\")\n",
      "sns.set_context(\"paper\", font_scale=1.8, rc={\"lines.linewidth\": 2})\n",
      "sns.kdeplot(delta32sim2, shade=True, color=\"#898989\",label='$\\delta_{sim2}$')\n",
      "sns.kdeplot(delta32sim3, shade=True, color=\"#262626\", label='$\\delta_{sim3}$')\n",
      "sns.axlabel('Log Likelihood ratio ($\\delta$)', 'Density')\n",
      "sns.despine()\n",
      "\n",
      "#plt.vlines(delta32star, 0, 0.061, lw=3, color=\"#262626\", \n",
      "#           linestyles='dashed', label=\"$\\delta^{*}$\")\n",
      "plt.vlines(delta32obs, 0, 0.061, lw=3, color='#F65655',\n",
      "           linestyles='dashed', label='$\\delta_{obs}$')\n",
      "plt.legend(loc=1)\n",
      "plt.title(\"(B) models 2 vs. 3\")\n",
      "plt.savefig(\"delta32.png\", dpi=300)\n",
      "plt.savefig(\"delta32.svg\")"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZQAAAEbCAYAAAD9I3KtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUlBCIYQQohT/Bw23caFVNX5AAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x237eb290>"
       ]
      }
     ],
     "prompt_number": 616
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Calculate % overlap\n",
      "Our power to reject model 3 is measured by the proportion of the simulations under model 2 that are to the left of the critical delta32star from data simulated under model 3. Remember this value is the 95 percentile of the model 3 bootstraps (0.05 significance), and at this level we find we have 100% power to reject model 3 when model 2 is true (no overlap). "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "1.-sum(delta32sim2 < delta32star)/float(len(delta32sim2))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 599,
       "text": [
        "1.0"
       ]
      }
     ],
     "prompt_number": 599
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## Now the same comparison against model 1"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Load bootstrap data simulated under model 1"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "W = ['title','N1','N2','N3','T2','T1','m13','m31','m23','m32','LL']\n",
      "LL11 = pd.read_table(\"dadi_m1_f1_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]\n",
      "\n",
      "W = ['title','N1','N2','N3','T2','T1','m13','m31','m23','m32','LL']\n",
      "LL12 = pd.read_table(\"dadi_m1_f2_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]\n",
      "\n",
      "W = ['title','N1','N2','N3','T2','T1','f','LL']\n",
      "LL13 = pd.read_table(\"dadi_m1_f3_boots.txt\", sep=\"\\t\", names=W, index_col=0).sort()[\"LL\"]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 558
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Calculate $\\delta_{31}$ under model 1 sims"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta31sim1 = -2.*(LL13-LL11)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 617
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Compare $\\delta_{31}$ simulated under both models"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.rcParams['figure.figsize'] = (6,4)\n",
      "plt.gcf().subplots_adjust(bottom=0.20, left=0.15)\n",
      "\n",
      "sns.set_style(\"ticks\")\n",
      "sns.set_context(\"paper\", font_scale=1.8, rc={\"lines.linewidth\": 2})\n",
      "sns.kdeplot(delta31sim1, shade=True, color=\"#898989\",label='$\\delta_{sim1}$')\n",
      "sns.kdeplot(delta31sim3, shade=True, color=\"#262626\", label='$\\delta_{sim3}$')\n",
      "sns.axlabel('Log Likelihood ratio ($\\delta$)', 'Density')\n",
      "sns.despine()\n",
      "\n",
      "plt.vlines(delta31obs, 0, 0.085, lw=3, color='#F65655',\n",
      "           linestyles='dashed', label='$\\delta_{obs}$')\n",
      "plt.legend(loc=1)\n",
      "plt.title(\"(A) models 1 vs. 3\")\n",
      "plt.savefig(\"delta31.png\", dpi=300)\n",
      "plt.savefig(\"delta31.svg\")"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZQAAAEbCAYAAAD9I3KtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x237c9290>"
       ]
      }
     ],
     "prompt_number": 618
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Calculate % overlap\n",
      "Our power to reject model 3 is measured by the proportion of the simulations under model 1 that are to the left of the critical delta31star from data simulated under model 3. Remember this value is the 95 percentile of the model 3 bootstraps (0.05 significance), and at this level we find we have 92% power to reject model 3 when model 1 is true (no overlap). "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "1.-sum(delta31sim1 < delta31star)/float(len(delta31sim1))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 619,
       "text": [
        "0.99004975124378114"
       ]
      }
     ],
     "prompt_number": 619
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "------------------   \n",
      "# Compare models 2 & 1"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "We'll begin by assuming model 1 is the true model. We can see already that the difference between models 2 and 1 is much smaller than it was between 2 and 3. But model 2 is a slightly better fit to the empirical data. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta12obs = -2.*(ML1[\"LL\"] - ML2[\"LL\"])\n",
      "delta13obs = -2.*(ML1[\"LL\"] - ML3[\"LL\"])\n",
      "\n",
      "print 'LL model3:', ML3[\"LL\"]\n",
      "print 'LL model2:', ML2[\"LL\"]\n",
      "print 'LL model1:', ML1[\"LL\"]\n",
      "print 'delta12 observed:', delta12obs\n",
      "print 'delta13 observed:', delta13obs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "LL model3: -555.343768211\n",
        "LL model2: -541.902480159\n",
        "LL model1: -543.081535912\n",
        "delta12 observed: 2.35811150515\n",
        "delta13 observed: -24.5244645982\n"
       ]
      }
     ],
     "prompt_number": 623
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### How much better?\n",
      "We compare the difference in their fit to the real data to the difference between them when data is simulated under model 1."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta12sim1 = -2.*(LL11-LL12)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 624
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Is our observed $\\delta$ > 95% of the bootstrap values simulated under model 1?\n",
      "Not quite, but we can reject model 1 in favor of model 2 at a P-value of 0.07."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print 'obs=', delta12obs\n",
      "delta12star = np.percentile(delta12sim1, 95)\n",
      "print '\\ndelta*95=',delta12star\n",
      "print delta12obs > delta12star\n",
      "delta12star = np.percentile(delta12sim1, 94)\n",
      "print '\\ndelta*94=',delta12star\n",
      "print delta12obs > delta12star\n",
      "delta12star = np.percentile(delta12sim1, 93)\n",
      "print '\\ndelta*93=',delta12star\n",
      "print delta12obs > delta12star"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "obs= 2.35811150515\n",
        "\n",
        "delta*95= 3.257462352\n",
        "False\n",
        "\n",
        "delta*94= 2.603696052\n",
        "False\n",
        "\n",
        "delta*93= 1.47879285\n",
        "True\n"
       ]
      }
     ],
     "prompt_number": 669
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "----------------  \n",
      "\n",
      "## Power to reject model 1\n",
      "What power do we have to reject model 1 if model 2 is the correct model?"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Calculate $\\delta_{12}$ under model 2 sims"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "delta12sim2 = -2.*(LL21-LL22)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 630
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Compare $\\delta_{12}$ simulated under both models"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.rcParams['figure.figsize'] = (6,4)\n",
      "plt.gcf().subplots_adjust(bottom=0.20, left=0.15)\n",
      "\n",
      "sns.set_context(\"paper\", font_scale=1.8, rc={\"lines.linewidth\": 2})\n",
      "sns.kdeplot(delta12sim1, shade=True, color=\"#262626\", label='$\\delta_{sim1}$')\n",
      "sns.kdeplot(delta12sim2, shade=True, color=\"#898989\", label='$\\delta_{sim2}$')\n",
      "\n",
      "sns.axlabel('Log Likelihood ratio ($\\delta$)', 'Density')\n",
      "sns.despine()\n",
      "\n",
      "plt.vlines(delta12obs, 0, 0.055, lw=3, color='#F65655',\n",
      "           linestyles='dashed', label='$\\delta_{obs}$')\n",
      "\n",
      "plt.legend(loc=2)\n",
      "plt.title(\"(C) models 1 vs. 2\")\n",
      "plt.savefig(\"delta12.png\", dpi=300)\n",
      "plt.savefig(\"delta12.svg\")"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZQAAAEbCAYAAAD9I3KtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsocUGTtYAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x22dcc850>"
       ]
      }
     ],
     "prompt_number": 631
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Calculate % overlap and power"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Our power to reject model 1 when model 2 at 0.05 false positive rate is only 85%, given our data. "
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "sum(delta12sim2 > delta12star)/float(len(delta12sim2))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 632,
       "text": [
        "0.85572139303482586"
       ]
      }
     ],
     "prompt_number": 632
    }
   ],
   "metadata": {}
  }
 ]
}