SIRD_Covid19_Colombia.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Modelo SIRD de la pandemia COVID-19 en Colombia\n",
    "\n",
    "Por:\n",
    "\n",
    "Diego Andrés Alvarez Marín, Phd\n",
    "\n",
    "Profesor asociado en dedicación exclusiva\n",
    "\n",
    "Departamento de Ingeniería Civil, Facultad de Ingeniería y Arquitectura\n",
    "\n",
    "Universidad Nacional de Colombia, Sede Manizales\n",
    "\n",
    "Correo electrónico: [daalvarez@unal.edu.co](daalvarez@unal.edu.co)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El objetivo de este cuaderno es realizar una proyección del COVID19 en Colombia basado en el modelo epidemiológico **SIRD** (**S**=susceptibles, **I**=infectados, **R**=recuperados, **D**=deaths - muertes). Los datos empleados en la modelación corresponden a los que han sido públicamente difundidos por el [Instituto Nacional  de  Salud](https://www.ins.gov.co).\n",
    "\n",
    "## El modelo SIRD\n",
    "El modelo SIRD es uno de los [modelos epidemiológicos](https://en.wikipedia.org/wiki/Compartmental_models_in_epidemiology) más sencillos y populares, ya que este es capaz de capturar muchas de las características típicas de los brotes epidémicos. Este modelo determinista es una generalización del llamado modelo SIR, que fue propuesto por [W. O. Kermack y A. G. McKendrick](https://en.wikipedia.org/wiki/Kermack%E2%80%93McKendrick_theory) en 1927.\n",
    "\n",
    "El modelo divide la población expuesta a la infección en cuatro grupos mutuamente excluyentes:\n",
    "\n",
    "* Población susceptible ($S$): individuos sin inmunidad al agente infeccioso y que por tanto pueden ser infectados si son expuestos a este.\n",
    "\n",
    "* Población infectada ($I$): individuos que están infectados en un momento dado y pueden transmitir la infección a personas de la población susceptibles con los que entran en contacto.\n",
    "\n",
    "* Población recuperada ($R$), individuos que se curaron de la infección, que ahora son inmunes y que ya no la propagan.\n",
    "\n",
    "* Defunciones/Fallecidos ($D$): individuos que murieron a causa del agente infeccioso.\n",
    "\n",
    "La suma de estas cuatro variables es la constante $N = S+I+R+D$ y es igual a la población total, que en Colombia equivale a:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "N = 50_372_424 # personas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El modelo SIRD relaciona los cambios en estas cuatro poblaciones a través tres parámetros:\n",
    "* la tasa de infección o transmisión $\\beta$ que representa el número promedio de individuos susceptibles que una infectado contagia por día. Esta variable depende del grado de distanciamiento social y de las prácticas higiénicas que adopte la población. Poco distanciamiento social y malas prácticas higiénicas se reflejan en un $\\beta$ alto.\n",
    "* la tasa de recuperación $\\gamma$, la cual representa la tasa de personas infectadas que se recuperan y dejan de ser contagiosas.\n",
    "* la tasa de mortalidad $\\mu$, la cual representa la tasa de personas infectadas que mueren.\n",
    "\n",
    "El modelo SIRD utiliza el sistema de ecuaciones diferenciales ordinarias no lineales:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$$\n",
    "\\begin{align}\n",
    "& \\frac{dS(t)}{dt} = - \\frac{\\beta I(t) S(t)}{N} \\\\\n",
    "& \\frac{dI(t)}{dt} = \\frac{\\beta I(t) S(t)}{N} - \\gamma I(t) - \\mu I(t) \\\\\n",
    "& \\frac{dR(t)}{dt} = \\gamma I(t) \\\\\n",
    "& \\frac{dD(t)}{dt} = \\mu I(t)\n",
    "\\end{align}\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "donde la variable tiempo $t$ se modela en días. Dicho sistema de ecuaciones diferenciales se puede modelar en el lenguaje de programación Python así:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# S(t): número de personas susceptibles en el día t\n",
    "# I(t): número de personas infectadas   en el día t\n",
    "# R(t): número de personas recuperadas  en el día t\n",
    "# D(t): número de personas fallecidas   en el día t\n",
    "def dx_dt_SIRD(x, t, param):\n",
    "    S, I, R, D = x\n",
    "    beta, gamma, mu = param\n",
    "    dS_dt = -beta*S*I/N\n",
    "    dI_dt =  beta*S*I/N - gamma*I - mu*I\n",
    "    dR_dt =  gamma*I\n",
    "    dD_dt =  mu*I\n",
    "    return [dS_dt, dI_dt, dR_dt, dD_dt]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Limitaciones y suposiciones del modelo SIRD\n",
    "A continuación se listan las suposiciones y limitaciones que tiene el modelo SIRD:\n",
    "* La interacción entre los individuos es aleatoria.\n",
    "* Las personas recuperadas no se reinfectan.\n",
    "* No se tiene en cuenta que la población $N$ es variable debido a fenómenos como nacimientos y muertes por otras causas y viajes al extranjero.\n",
    "* La población es de tamaño fijo. El modelo no tiene en cuenta que después de infectada una persona existe un periodo en el cual el infectado no propaga la enfermedad. Para el COVID19 se dice que el periodo de propagación por parte de una persona infectada empieza aproximadamente 3 días antes de presentar síntomas visibles y se mantiene por 7 a 9 días después de este evento.\n",
    "* Las políticas públicas que propenden por el control de la infección no cambian en el tiempo y están representadas en la muestra inicial con la que se ajusta el modelo.\n",
    "* La naturaleza infecciosa del virus no cambia en el tiempo; en otras palabras, el modelo no tiene en cuenta que las enfermedades tienen periodos en las cuales son más contagiosas.\n",
    "* El modelo no tiene en cuenta que la población está distribuída de forma heterogénea por el país, ni que sus factores socioeconómicos son diferentes.\n",
    "* De acuerdo a las ecuaciones diferenciales el término $\\frac{\\beta I(t) S(t)}{N}$ refleja el hecho de que todos los infectados tienen igual capacidad de contagiar el virus y que todas las personas susceptibles tienen las mismas probabilidades de contagiarse, lo cual es poco ideal en un escenario real. Por lo tanto, el modelo asume que la población es homogénea.\n",
    "* El modelo no tiene en cuenta que al saturarse el sistema de salud, la tasa de mortalidad $\\mu$ tiende a crecer."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Datos para el ajuste"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Se utilizaron los datos que son publicados diariamente por el [Instituto Nacional  de  Salud](https://www.ins.gov.co/Noticias/paginas/coronavirus.aspx). En particular, se usó la base de datos tabulada por [ESRI Colombia](http://datosabiertos.esri.co/datasets). De dicha información se extraen los datos a partir de abril 10 de 2020, ya que en abril 5 y 6 hay una inconsistencia en los datos (no suman correctamente) y en abril 8 y 9 las fechas fueron presentadas con un formato diferente al resto, lo que podría corresponder a un error. De este modo, la base de datos útil se convierte en:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>TOTAL_CASOS</th>\n",
       "      <th>TOTAL_MUERTES</th>\n",
       "      <th>TOTAL_RECUPERADOS</th>\n",
       "      <th>FECHA_ACTUALIZACION</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2473</td>\n",
       "      <td>80</td>\n",
       "      <td>197</td>\n",
       "      <td>2020/04/10 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2709</td>\n",
       "      <td>100</td>\n",
       "      <td>214</td>\n",
       "      <td>2020/04/11 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2776</td>\n",
       "      <td>109</td>\n",
       "      <td>270</td>\n",
       "      <td>2020/04/12 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2852</td>\n",
       "      <td>112</td>\n",
       "      <td>319</td>\n",
       "      <td>2020/04/13 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2979</td>\n",
       "      <td>127</td>\n",
       "      <td>354</td>\n",
       "      <td>2020/04/14 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>327850</td>\n",
       "      <td>11017</td>\n",
       "      <td>173727</td>\n",
       "      <td>2020/08/03 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>334979</td>\n",
       "      <td>11315</td>\n",
       "      <td>180258</td>\n",
       "      <td>2020/08/04 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>345714</td>\n",
       "      <td>11624</td>\n",
       "      <td>186317</td>\n",
       "      <td>2020/08/05 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>357710</td>\n",
       "      <td>11939</td>\n",
       "      <td>192355</td>\n",
       "      <td>2020/08/06 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>367196</td>\n",
       "      <td>12250</td>\n",
       "      <td>198495</td>\n",
       "      <td>2020/08/07 05:00:00+00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>120 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     TOTAL_CASOS  TOTAL_MUERTES  TOTAL_RECUPERADOS     FECHA_ACTUALIZACION\n",
       "0           2473             80                197  2020/04/10 05:00:00+00\n",
       "1           2709            100                214  2020/04/11 05:00:00+00\n",
       "2           2776            109                270  2020/04/12 05:00:00+00\n",
       "3           2852            112                319  2020/04/13 05:00:00+00\n",
       "4           2979            127                354  2020/04/14 05:00:00+00\n",
       "..           ...            ...                ...                     ...\n",
       "115       327850          11017             173727  2020/08/03 05:00:00+00\n",
       "116       334979          11315             180258  2020/08/04 05:00:00+00\n",
       "117       345714          11624             186317  2020/08/05 05:00:00+00\n",
       "118       357710          11939             192355  2020/08/06 05:00:00+00\n",
       "119       367196          12250             198495  2020/08/07 05:00:00+00\n",
       "\n",
       "[120 rows x 4 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import locale\n",
    "locale.setlocale(locale.LC_ALL,'es_CO.utf8')\n",
    "\n",
    "import pandas as pd\n",
    "dfcompleto = pd.read_csv('Colombia_COVID19_Coronavirus_casos_diarios.csv', \n",
    "      skiprows=[i for i in range(1,36)],\n",
    "      usecols=['FECHA_ACTUALIZACION','TOTAL_RECUPERADOS', 'TOTAL_MUERTES', 'TOTAL_CASOS'])\n",
    "\n",
    "dfcompleto"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Para hacer la regresión, se seleccionaron los últimos 21 días,"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# La regresión se hará utilizando los últimos \"Ndata\" días de la base de datos\n",
    "Ndata = 21 # días = 3 semanas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "ya que la información más reciente representa las acciones gubernamentales en curso y la historia de contagios, recuperación y muerte en el país:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Utilizar los datos de los últimos \"Ndata\" días para hacer el ajuste\n",
    "df = dfcompleto.tail(Ndata) \n",
    "fecha_regresion = pd.to_datetime(df['FECHA_ACTUALIZACION']).dt.date\n",
    "fecha_hoy = fecha_regresion.iloc[-1]\n",
    "Rd = df['TOTAL_RECUPERADOS'].to_numpy() # número de recuperados\n",
    "Dd = df['TOTAL_MUERTES'].to_numpy()     # número de fallecidos\n",
    "Td = df['TOTAL_CASOS'].to_numpy()       # número de casos reportados\n",
    "Id = Td - Dd - Rd                       # número de infectados\n",
    "Sd = N - Td                             # número de personas susceptibles"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "de modo que la base de datos empleada para hacer la regresión se convierte en:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>TOTAL_CASOS</th>\n",
       "      <th>TOTAL_MUERTES</th>\n",
       "      <th>TOTAL_RECUPERADOS</th>\n",
       "      <th>FECHA_ACTUALIZACION</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>190700</td>\n",
       "      <td>6516</td>\n",
       "      <td>85836</td>\n",
       "      <td>2020/07/18 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>197278</td>\n",
       "      <td>6736</td>\n",
       "      <td>91793</td>\n",
       "      <td>2020/07/19 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>204005</td>\n",
       "      <td>6929</td>\n",
       "      <td>95804</td>\n",
       "      <td>2020/07/20 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>211038</td>\n",
       "      <td>7166</td>\n",
       "      <td>98840</td>\n",
       "      <td>2020/07/21 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>218428</td>\n",
       "      <td>7373</td>\n",
       "      <td>101613</td>\n",
       "      <td>2020/07/22 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>226373</td>\n",
       "      <td>7688</td>\n",
       "      <td>107951</td>\n",
       "      <td>2020/07/23 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>233541</td>\n",
       "      <td>7975</td>\n",
       "      <td>113864</td>\n",
       "      <td>2020/07/24 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>240795</td>\n",
       "      <td>8269</td>\n",
       "      <td>119667</td>\n",
       "      <td>2020/07/25 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>248976</td>\n",
       "      <td>8525</td>\n",
       "      <td>125037</td>\n",
       "      <td>2020/07/26 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>257101</td>\n",
       "      <td>8777</td>\n",
       "      <td>131161</td>\n",
       "      <td>2020/07/27 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>267385</td>\n",
       "      <td>9074</td>\n",
       "      <td>136690</td>\n",
       "      <td>2020/07/28 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>276055</td>\n",
       "      <td>9454</td>\n",
       "      <td>142777</td>\n",
       "      <td>2020/07/29 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>286020</td>\n",
       "      <td>9810</td>\n",
       "      <td>148695</td>\n",
       "      <td>2020/07/30 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>295508</td>\n",
       "      <td>10105</td>\n",
       "      <td>154387</td>\n",
       "      <td>2020/07/31 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113</th>\n",
       "      <td>306181</td>\n",
       "      <td>10330</td>\n",
       "      <td>160708</td>\n",
       "      <td>2020/08/01 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>317651</td>\n",
       "      <td>10650</td>\n",
       "      <td>167239</td>\n",
       "      <td>2020/08/02 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>327850</td>\n",
       "      <td>11017</td>\n",
       "      <td>173727</td>\n",
       "      <td>2020/08/03 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>334979</td>\n",
       "      <td>11315</td>\n",
       "      <td>180258</td>\n",
       "      <td>2020/08/04 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>345714</td>\n",
       "      <td>11624</td>\n",
       "      <td>186317</td>\n",
       "      <td>2020/08/05 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>357710</td>\n",
       "      <td>11939</td>\n",
       "      <td>192355</td>\n",
       "      <td>2020/08/06 05:00:00+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>367196</td>\n",
       "      <td>12250</td>\n",
       "      <td>198495</td>\n",
       "      <td>2020/08/07 05:00:00+00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     TOTAL_CASOS  TOTAL_MUERTES  TOTAL_RECUPERADOS     FECHA_ACTUALIZACION\n",
       "99        190700           6516              85836  2020/07/18 05:00:00+00\n",
       "100       197278           6736              91793  2020/07/19 05:00:00+00\n",
       "101       204005           6929              95804  2020/07/20 05:00:00+00\n",
       "102       211038           7166              98840  2020/07/21 05:00:00+00\n",
       "103       218428           7373             101613  2020/07/22 05:00:00+00\n",
       "104       226373           7688             107951  2020/07/23 05:00:00+00\n",
       "105       233541           7975             113864  2020/07/24 05:00:00+00\n",
       "106       240795           8269             119667  2020/07/25 05:00:00+00\n",
       "107       248976           8525             125037  2020/07/26 05:00:00+00\n",
       "108       257101           8777             131161  2020/07/27 05:00:00+00\n",
       "109       267385           9074             136690  2020/07/28 05:00:00+00\n",
       "110       276055           9454             142777  2020/07/29 05:00:00+00\n",
       "111       286020           9810             148695  2020/07/30 05:00:00+00\n",
       "112       295508          10105             154387  2020/07/31 05:00:00+00\n",
       "113       306181          10330             160708  2020/08/01 05:00:00+00\n",
       "114       317651          10650             167239  2020/08/02 05:00:00+00\n",
       "115       327850          11017             173727  2020/08/03 05:00:00+00\n",
       "116       334979          11315             180258  2020/08/04 05:00:00+00\n",
       "117       345714          11624             186317  2020/08/05 05:00:00+00\n",
       "118       357710          11939             192355  2020/08/06 05:00:00+00\n",
       "119       367196          12250             198495  2020/08/07 05:00:00+00"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preparación del modelo\n",
    "Para la solución del modelo matemático, requiere la especificación de unos valores tentativos de los parámetros y unas condiciones iniciales. \n",
    "\n",
    "### Parámetros iniciales\n",
    "Para la especificación de los valores tentativos de los parámetros iniciales, se consultó en la [literatura](https://arxiv.org/pdf/2003.14288.pdf) los rangos de valores asociadas a las constantes $\\beta$, $\\gamma$ y $\\mu$. De esta forma se especificaron:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "beta0  = 0.044      # tasa de transmisión  [1/día]\n",
    "gamma0 = 0.015      # tasa de recuperación [1/día]\n",
    "mu0    = 0.0045     # tasa de mortalidad   [1/día]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Condiciones iniciales\n",
    "Se definieron como condiciones iniciales aquellas asociadas al primer día del rango de datos usado para la regresión:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %% Se definen las condiciones iniciales de la ecuación diferencial\n",
    "x0 = [Sd[0], Id[0], Rd[0], Dd[0]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Los cuales corresponden a la fecha:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'julio 18 de 2020'"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.to_datetime(df['FECHA_ACTUALIZACION']).dt.date.iloc[0].strftime('%B %d de %Y')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regresión no lineal\n",
    "Con la información anterior se realizó una regresión no lineal para ajustar los datos colombianos al modelo matemático:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from scipy.integrate import odeint\n",
    "from scipy.optimize import least_squares\n",
    "\n",
    "# Eje del tiempo [días] para los datos de la regresión\n",
    "t = np.arange(Ndata) \n",
    "\n",
    "# Se encuentran los parámetros del sistema de ecuaciones diferenciales que \n",
    "# mejor se ajusten a los datos \n",
    "def fun_residuo(params):\n",
    "    sol = odeint(dx_dt_SIRD, x0, t, args=(params, ))\n",
    "    S, I, R, D = sol.T\n",
    "    return np.array([ (S-Sd)/Sd[-1], (I-Id)/Id[-1], (R-Rd)/Rd[-1], (D-Dd)/Dd[-1] ]).flatten()\n",
    "\n",
    "paramsLS = least_squares(fun_residuo, [beta0, gamma0, mu0])\n",
    "beta, gamma, mu = paramsLS.x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Obteniendo los siguientes resultados:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Los parámetros encontrados son:\n",
      "beta  = 0.071208\n",
      "gamma = 0.045882\n",
      "mu    = 0.002344\n"
     ]
    }
   ],
   "source": [
    "print('Los parámetros encontrados son:')\n",
    "print(f'beta  = {beta:.6f}')\n",
    "print(f'gamma = {gamma:.6f}')\n",
    "print(f'mu    = {mu:.6f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "De ellos se puede obtener dos variables relevantes:\n",
    "* El *número básico de reproducción* $R_0 = \\frac{\\beta}{\\gamma+\\mu}$, el cual indica la cantidad de personas que en promedio contagiaría un infectado durante su enfermedad. Este número es bastante importante, ya que cuando $R_0 < 1$, por un tiempo prolongado, se dice que la pandemia está controlada."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Número básico de reproducción R_0 = 1.48\n"
     ]
    }
   ],
   "source": [
    "print(f'Número básico de reproducción R_0 = {beta/(gamma+mu):.2f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* El *período infeccioso medio* $\\frac{1}{\\gamma+\\mu}$ que representa el tiempo promedio (en días) que un infectado es diseminador de la infección."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Periodo infeccioso promedio = 20.74 días\n"
     ]
    }
   ],
   "source": [
    "print(f'Periodo infeccioso promedio = {1/(gamma+mu):.2f} días')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Estimación de la evolución del COVID 19 en Colombia\n",
    "Una vez encontrados los parámetros del modelo se hizo la estimación de la evolución del COVID 19 para los siguientes 15 meses utilizando el siguiente código:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %% Resuelva la ecuación diferencial utilizando los parámetros encontrados\n",
    "N_dias_pronostico = Ndata + 15*30 # días\n",
    "\n",
    "t = np.arange(N_dias_pronostico)\n",
    "sol = odeint(dx_dt_SIRD, x0, t, args=(paramsLS.x, ))\n",
    "S, I, R, D = sol.T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La calidad de la regresión se puede apreciar en el siguiente gráfico, el cual muestra el buen ajuste del modelo matemático a la muestra:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "plt.rcParams[\"figure.figsize\"] = 13,8\n",
    "\n",
    "esc = 1_000 # miles\n",
    "df_res = pd.DataFrame({\n",
    "           'fecha':fecha_regresion, \n",
    "           #'Estimación Susceptibles':S[:Ndata]/esc,\n",
    "           'Estimación Casos confirmados':(I[:Ndata]+R[:Ndata]+D[:Ndata])/esc,\n",
    "           'Estimación Infectados':I[:Ndata]/esc,\n",
    "           'Estimación Recuperados':R[:Ndata]/esc, \n",
    "           'Estimación Decesos':D[:Ndata]/esc,\n",
    "           #'Dato Susceptibles':Sd/esc, \n",
    "           'Dato Casos confirmados':Td/esc, \n",
    "           'Dato Infectados':Id/esc, \n",
    "           'Dato Recuperados':Rd/esc, \n",
    "           'Dato Decesos':Dd/esc, \n",
    "         }).set_index('fecha')\n",
    "fig, ax = plt.subplots()\n",
    "df_res.plot(style=['m-','b-','r-','g-','m.','b.','r.','g.'], ax=ax)\n",
    "ax.set_xlabel(\"Fecha\")\n",
    "ax.set_ylabel(\"Personas [miles]\")\n",
    "ax.set_title(f'Ajuste del modelo SIRD a los datos de los últimos {Ndata} días\\n'\n",
    "             f'Fecha de ajuste: {fecha_hoy.strftime(\"%b %d, %Y\")}')\n",
    "ax.legend(loc='best')\n",
    "fig.autofmt_xdate() # rote fechas para que se vean mejor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Y haciendo una proyección sobre los siguientes 15 meses resulta:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# %% Gráfico del pronóstico\n",
    "esc = 1_000_000 # millones\n",
    "fecha_pronostico = pd.date_range(start=fecha_regresion.iloc[0], periods=N_dias_pronostico)\n",
    "df_res = pd.DataFrame({\n",
    "                   'fecha':fecha_pronostico, \n",
    "                   'Susceptibles':S/esc,                     \n",
    "                   'Infectados':I/esc,\n",
    "                   'Recuperados':R/esc, \n",
    "                   'Decesos':D/esc,\n",
    "                   'Casos confirmados':(I + R + D)/esc,                   \n",
    "                  }).set_index('fecha')\n",
    "fig, ax = plt.subplots()\n",
    "df_res.plot(ax=ax)\n",
    "ax.set_xlabel(\"Fecha\")\n",
    "ax.set_ylabel(\"Personas [millones]\")\n",
    "ax.set_title('Pronóstico de la evolución del COVID 19 para Colombia\\n(modelo SIRD)') \n",
    "ax.legend(loc='best')\n",
    "ax.grid(True, which=\"both\")    \n",
    "ax.annotate(fecha_hoy.strftime('%b %d, %Y'), xy=(fecha_hoy,0.5), xytext=(fecha_hoy, 30),\n",
    "            arrowprops=dict(facecolor='black', shrink=0.05), rotation=90\n",
    "            )\n",
    "ax.autoscale(enable=True, axis='x', tight=True)\n",
    "\n",
    "# se grafica el eje de porcentaje\n",
    "Poblac_a_Porcent = lambda pob: 100*pob/(N/esc)\n",
    "Porcent_a_Poblac = lambda por: (por/100)*(N/esc)\n",
    "eje_porcentaje = ax.secondary_yaxis('right', functions=(Poblac_a_Porcent, Porcent_a_Poblac))\n",
    "eje_porcentaje.set_ylabel('Proporción de la población [%]')\n",
    "\n",
    "# plt.yscale('log')\n",
    "\n",
    "df_res.to_excel(\"pronosticos_SIRD.xlsx\")\n",
    "plt.savefig('pronostico_COVID.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "De esta gráfica se observa que el pico de la pandemia estará alrededor de:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'febrero 18 de 2021'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.to_datetime(df_res['Infectados'].idxmax()).strftime('%B %d de %Y')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Que al final de la misma tendremos aproximadamente:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Número de defunciones al final de la pandemia = 1378693 personas\n"
     ]
    }
   ],
   "source": [
    "print(f'Número de defunciones al final de la pandemia = {D[-1]:.0f} personas')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Y, según [esta referencia](https://www.eafit.edu.co/escuelas/economiayfinanzas/cief/Documents/informe-especial-2020-abril-2.pdf), que la llamada *inmunidad colectiva* o *inmunidad de rebaño* comenzaría en:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'mayo 17 de 2021'"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.to_datetime((df_res['Susceptibles']-df_res['Casos confirmados']).abs().idxmin()).strftime('%B %d de %Y')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " esto es, cuando la curva de los susceptibles intercepte la curva de los casos confirmados."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Análisis de resultados y comentarios finales"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Los resultados aquí mostrados nos dan una pista de que el pico de la pandemia se presentará en el primer trimestre de 2021 y terminará en el segundo semestre del 2021. De todas formas, dichas fechas tentativas dependerán mucho de las acciones gubernamentales adoptadas, de la cultura ciudadana de autocuidado y distanciamiento social y del avance en las investigaciones de vacunas, fármacos y tratamientos para el manejo del COVID.\n",
    "\n",
    "El modelo utilizado es muy sensible a las condiciones iniciales, a los parámetros $\\beta$, $\\gamma$ y $\\mu$ y a la calidad de los datos epidemiológicos empleados; además, según lo mencionado en la sección ***Limitaciones y suposiciones del modelo SIRD***, son muchos los factores que intervienen en la dinámica de la pandemia que no están apropiadamente representadas en el modelo matemático, el cual puede llevar a variaciones en los resultados.\n",
    "\n",
    "El modelo muestra que es nuestra responsabilidad adoptar medidas personales de aislamiento y buena higiene, ya que de lo contrario más de un millón de personas podrían fallecer.\n",
    "\n",
    "Finalmente, tenga en cuenta que yo no soy un epidemiólogo; soy un ingeniero que aplica estas técnicas en la modelación de estructuras. Sería de mi agrado que epidemiólogos, virólogos y salubristas y otros profesionales de la salud me hicieran comentarios al respecto de este análisis."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}