diff --git a/tutorials/units-and-integration/units-and-integration.ipynb b/tutorials/units-and-integration/units-and-integration.ipynb
old mode 100755
new mode 100644
index 18606d5e..32f461c4
--- a/tutorials/units-and-integration/units-and-integration.ipynb
+++ b/tutorials/units-and-integration/units-and-integration.ipynb
@@ -1,366 +1,550 @@
{
- "cells": [
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "# Using `scipy.integrate`\n",
- "\n",
- "## Authors\n",
- "Zach Pace, Lia Corrales, Stephanie T. Douglas\n",
- "\n",
- "## Learning Goals\n",
- "* perform numerical integration in the `astropy` and scientific python context\n",
- " * trapezoidal approximation\n",
- " * gaussian quadrature\n",
- "* use `astropy`'s built-in black-body curves\n",
- "* understand how `astropy`'s units interact with one another\n",
- "* define a Python class\n",
- " * how the `__call__` method works\n",
- "* add $\\LaTeX$ labels to `matplotlib` figures using the `latex_inline` formatter\n",
- "\n",
- "## Keywords\n",
- "modeling, units, synphot, OOP, LaTeX, astrostatistics, matplotlib, units, physics\n",
- "\n",
- "## Companion Content\n",
- "* http://synphot.readthedocs.io/en/latest/\n",
- "* [Using Astropy Quantities for astrophysical calculations](http://www.astropy.org/astropy-tutorials/rst-tutorials/quantities.html)\n",
- "\n",
- "## Summary\n",
- "In this tutorial, we will use the examples of the Planck function and the stellar initial mass function (IMF) to illustrate how to integrate numerically, using the trapezoidal approximation and Gaussian quadrature. We will also explore making a custom class, an instance of which is callable in the same way as a function. In addition, we will encounter `astropy`'s built-in units, and get a first taste of how to convert between them. Finally, we will use $\\LaTeX$ to make our figure axis labels easy to read."
- ]
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "WIVPH0GmXNbD"
+ },
+ "source": [
+ "# Usando `scipy.integrate`\n",
+ "\n",
+ "## Autores\n",
+ "Zach Pace, Lia Corrales, Stephanie T. Douglas\n",
+ "\n",
+ "## Tradução\n",
+ "\n",
+ "Luciano Cordeiro\n",
+ "\n",
+ "## Objetivo do aprendizado\n",
+ "* realizar integração numérica no contexto `astropy` e científico python\n",
+ " * aproximação trapezoidal\n",
+ " * quadratura gaussiana\n",
+ "* usar as curvas de corpo negro embutidas do `astropy`\n",
+ "* entender como as unidades do `astropy` interagem umas com as outras\n",
+ "* definir uma classe Python\n",
+ " * como funciona o método `__call__`\n",
+ "* adicionar rótulos $\\LaTeX$ em figuras `matplotlib` usuando o formatador `latex_inline` \n",
+ "\n",
+ "## Palavras-chaves \n",
+ "modelagem, unidades, synphot, OOP, LaTeX, astroestatística, matplotlib, física\n",
+ "\n",
+ "## Conteúdo complementar\n",
+ "* http://synphot.readthedocs.io/en/latest/\n",
+ "* [Using Astropy Quantities for astrophysical calculations](http://www.astropy.org/astropy-tutorials/rst-tutorials/quantities.html)\n",
+ "\n",
+ "## Sumário\n",
+ "Neste tutorial, usaremos os exemplos da função de Planck e da função de massa inicial estelar (IMF) para ilustrar como integrar numericamente, usando a aproximação trapezoidal e a quadratura gaussiana. Também exploraremos a criação de uma classe personalizada, cuja instância pode ser chamada da mesma maneira que uma função. Além disso, encontraremos as unidades integradas do astropy e teremos uma primeira amostra de como converter entre elas. Finalmente, usaremos o $\\LaTeX$ para facilitar a leitura dos rótulos dos eixos das figuras."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 10,
+ "metadata": {
+ "id": "RAdawwnRXNbH"
+ },
+ "outputs": [],
+ "source": [
+ "import numpy as np\n",
+ "from scipy import integrate\n",
+ "from astropy.modeling.models import BlackBody\n",
+ "from astropy import units as u, constants as c\n",
+ "import matplotlib.pyplot as plt\n",
+ "\n",
+ "%matplotlib inline"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "OhPAWSAlXNbJ"
+ },
+ "source": [
+ "## A função de Planck\n",
+ "\n",
+ "A função de Planck descreve como um corpo negro irradia energia. Vamos explorar como encontrar a luminosidade bolométrica usando a função de Planck tanto no espaço de frequência quanto no espaço de comprimento de onda. "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "aqXb3VB3XNbJ"
+ },
+ "source": [
+ "Digamos que temos um corpo negro a 5000 Kelvin. Podemos descobrir a intensidade total (bolométrica) deste objeto, integrando a função de Planck. \n",
+ "A maneira mais simples de fazer isso é aproximar a integral usando a regra do trapézio.\n",
+ "\n",
+ "Vamos fazer isso primeiro usando a definição de frequência da função de Planck. Vamos definir uma grade de frequência de fótons e avaliar a função de Planck nessas frequências. \n",
+ " \n",
+ "Esses serão usados para integrar numericamente usando a regra trapezoidal. Ao multiplicar uma matriz `numpy` por uma unidade `astropy`, obtemos uma `quantidade`, que é efetivamente uma combinação de um ou mais números e uma unidade."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "j07CMpw6XNbK"
+ },
+ "source": [
+ "
\n",
+ "**Nota sobre unidades de impressão**: \n",
+ "\n",
+ "Quantidades e unidades podem ser impressas em strings usando [Format String Syntax](https://docs.python.org/3/library/string.html#format-string-syntax). Esta demonstração usa o formato `latex_inline` que está embutido no pacote 'astropy.units`. Para ver formas adicionais de formatar quantidades, consulte [Getting Started](http://docs.astropy.org/en/stable/units/#getting-started) nas páginas de documentação do astropy.units.\n",
+ "
"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 299
+ },
+ "id": "yJD8dBdYXNbL",
+ "outputId": "3f8f436e-d17e-4315-bcdf-cb743d3e6718"
+ },
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ }
+ }
+ ],
+ "source": [
+ "bb = BlackBody(5000. * u.Kelvin)\n",
+ "\n",
+ "nu = np.linspace(1., 3000., 1000) * u.THz\n",
+ "bb5000K_nu = bb(nu)\n",
+ "plt.plot(nu, bb5000K_nu)\n",
+ "plt.xlabel(r'$\\nu$, [{0:latex_inline}]'.format(nu.unit))\n",
+ "plt.ylabel(r'$I_{\\nu}$, ' + '[{0:latex_inline}]'.format(bb5000K_nu.unit))\n",
+ "plt.title('Planck function in frequency')\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "pw9HsyXWXNbL"
+ },
+ "source": [
+ "### Usando $LaTeX$ para rotular eixos\n",
+ "\n",
+ "Aqui, usamos a marcação $LaTeX$ para adicionar rótulos nos eixos com melhor aparência. Para fazer isso, colocamos o texto de marcação LaTeX em cifrões, dentro de uma string `r'\\$ ... \\$'`. O `r` antes das aspas indica que a string é \"raw\" e as barras invertidas são tratadas literalmente. Este é o formato sugerido para o texto do rótulo do eixo que inclui marcação. \n",
+ "\n",
+ " Agora integramos numericamente usando a regra do trapézio.\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 38
+ },
+ "id": "anjDzqDwXNbM",
+ "outputId": "d90bf5d4-3ae8-41d9-d788-e7700b63ce86"
+ },
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/latex": "$1.1280849 \\times 10^{10} \\; \\mathrm{\\frac{erg}{s\\,sr\\,cm^{2}}}$",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "execution_count": 3
+ }
+ ],
+ "source": [
+ "np.trapz(x=nu, y=bb5000K_nu).to('erg s-1 cm-2 sr-1')"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "M9xbbIV6XNbM"
+ },
+ "source": [
+ "Agora podemos fazer algo semelhante, mas para uma grade de comprimento de onda. Queremos integrar em uma faixa de comprimento de onda equivalente à faixa de frequência que fizemos anteriormente. Podemos transformar a frequência máxima no comprimento de onda (mínimo) correspondente usando o método `.to() `, com a adição de uma *equivalência*."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 303
+ },
+ "id": "dkNoQ4E0XNbN",
+ "outputId": "17627d50-cd87-4cfb-d282-95bcdea27e97"
+ },
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZkAAAEeCAYAAABYEGiuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5xUd73/8dd7tgNL2V0glIUlQAqphIUUEo0xmmJMLGnEbjS/eM1VrxqjV6/12u/Va2I0xnKtId2IisaouTGRECCBBEIaZRcWCGXou2z//P44Z2BYli2zM3tmZz/Px2Mee9qc85mz5bPfcr5fmRnOOedcJsSiDsA551zu8iTjnHMuYzzJOOecyxhPMs455zLGk4xzzrmM8STjnHMuYzzJuG5J+j9JH8iWa0g6XtIKSfskfSSTcXW47iRJ+yXlZeDc+yUdm+7zZpqk8yXVRXTtL0r6dRTXdj3nScYBIKlG0oHwj91WST+XNCzquI7iU8CjZlZqZrdm6iLhPbkwsW5mG8xsmJm1pfta4XnXpfu8uSLKZOb6xpOMS/ZmMxsGnAFUA5+LOJ6jmQw8H3UQzrnueZJxRzCzTcCfgJM77pM0VdLfJcUl7ZD0G0kjk/bXSPqkpOck7ZF0j6TipP1XhFVdeyWtlXRxJ9cYF77/5k72/R14HfD9sNR1XMeqNknvlfRE0rpJulHSK5J2S7pdkpL2f1DSC2H122pJZ0j6FTAJ+H14nU9JqgrPlR++b7ykBZJ2Sloj6YNJ5/yipHsl/TI87/OSqo92z8PzTguXfx7G+MfwvU9JmnqU9/1C0ifC5QnheT6c9L3aKSkmaZSkP0jaLmlXuDwxPO4aScs6nPffJC0Il4sk/ZekDWEp9w5JJUeJZ7ykB8LrrE+uzuzunoT3fXm4777wZ+c/JQ0l+HkcH34v9ksaH76tsKf32EXDk4w7gqRK4FJgeWe7ga8D44ETgUrgix2OuRq4GJgCnAq8NzzvHOCXwM3ASOA1QE2Ha08BHgO+b2bf7nhxM7sAeBy4KaxiermHH+syYHYYz9XAReH1rgrjfzcwHLgciJvZu4ANhKU7M/tWJ+e8G6gL78WVwNckXZC0//LwmJHAAuD7PYwV4FrgS8AoYA3w1aMc9xhwfrj8WmAdwX1NrD9uZu0Ev+v/S1AKnAQcSIrn98DxkqYnnfc64K5w+RvAccDpwDRgAvD5joFIioXnejY85vXAxyRdlHRYp/dEUiHwW+DnQBkwH3grgJnVA5cAm8PvxTAz29zV+Vz2yMkkI+lnkrZJWtXD468O/4N9XtJd3b8jZz0kaTfwBMEfr691PMDM1pjZI2bWZGbbge8Q/DFLdquZbTaznQR/dE4Pt18P/Cx8f7uZbTKzF5PeNwN4FPiCmd2Z5s/2DTPbbWYbwmskYvoA8C0zW2qBNWZW293JwkQ8F7jFzBrNbAXwE4JklfCEmS0M23B+BZzWi3h/a2ZLzKwV+E1SvB09Bpwb/oF/DfCtMC4Ivi+PAZhZ3MweMLMGM9tHkLReG+5rAH4HzAs/23TgBGBBWOK7Afg3M9sZvvdrBEmwo9nAaDP7spk1h21MP+5w7NHuyVlAPsHPTouZPQgs6cF96ss9dv0gJ5MMwX9DR1TDdCb8hfoMMNfMTgI+lsG4st1bzGykmU02s38xswMdD5A0VtLdkjZJ2gv8GqjocNirScsNQKIDQSWwtovrvwPYBNyf+kc4qlRjOprxQOKPbkItwX/wR7tmcaKqrQeOFu9hzGwtUE+QhM4D/gBslnQ8SUlG0hBJP5JUG37f/gGM1KGecncRJhmCUsxDYfIZDQwBng6rGncDfw63dzSZoEprd9Kx/w6M7eJzJe7JeGCTHT5i78aj3ZwenM9liZxMMmb2D2Bn8rawfvrPkp6W9LikE8JdHwRuN7Nd4Xu39XO4A83XAANOMbPhwDsJqtB6YiPQadtC6IvADuAu9a6bcD3BH8KEY3rx3q5i6mqI8s1AmaTSpG2TCJJkf3uMoLquMGxPewx4D0FV24rwmE8AxwNnht+3RJVa4nv3CDBa0ukEySZRot9BULV2UvgPyEgzGxF2EOloI7A+6biRYQ/AS3vwGbYAE8KSU0Jl0rIPFz9A5WSSOYo7gX81s1nAJ4EfhNuPA46T9E9Ji9VJQ7Q7TCmwH9gjaQJB+0pP/RR4n6TXh43RE5KSPUALcBUwFPhlWAXUEyuAt4X/rU8jqJbrqZ8An5Q0S4FpkiaH+7YCnT67YmYbgUXA1yUVSzo1vG4Uz208BtxEUDoB+L9w/Ymk7talBMlit6Qy4AvJJzCzFuA+4NsEbSKPhNvbCaq8vitpDBzsYJDczpKwBNgn6RZJJZLyJJ0saXYPPsOTQBtwk6R8SVcAc5L2bwXKJY3owblcFhkUSUbB8x7nAPdJWgH8CBgX7s4HphM0ns4Dfqyk3lLuCF8i6OK8B/gj8GBP32hmS4D3Ad8N3/8YQRVL8jHNwNsIqlh+1sNE812gmeAP0S8I2jB6GtN9BO0TdwH7gIcI/shC0MHhc2HVzyc7efs8oIqgVPNbgrakv/b02mn0GEESSSSZJwhKdv9IOuZ/gBKCksligiqvju4CLgTuC9uCEm4h6HywOKxq+ytBqegwYUK7jKDqbn14rZ8A3SaGpO/79cBughLyH4CmcP+LBJ0B1oXfj/FHO5fLLsrVScskVQF/MLOTJQ0HXjKzcZ0cdwfwlJn9b7j+N+DTZra0P+N1zh1O0lPAHYnfTTcwDYqSjJntBdaH3VUJq0USvVAeIuwCKqmCoPrMn7x2rp9Jeq2kY8LqsvcQdDfvrMTlBpCcTDKS5hPU8R4vqU7S9QQ9l66X9CzB0+JXhIc/DMQlrSbo2nqzmcWjiNu5Qe54gmdsdhN0VLjSzLZEG5Lrq5ytLnPOORe9nCzJOOecyw6eZJxzzmVMzj0ZW1FRYVVVVVGH4ZxzA8bTTz+9w8w6G8Whz3IuyVRVVbFs2bLuD3TOOQeApG7H60uVV5c555zLmMiSTHcjJUt6h4I5RVZKWpT0XItzzrkBIsqSzM/peqTk9cBrzewU4CsEY48555wbQCJrkzGzf4RDvxxt/6Kk1cXAxEzH5JxzLr0GSpvM9QTTrzrnnBtAsr53maTXESSZc7s45gaC2fuYNGlSP0XmnHOuO1ldkgnn6PgJcEVX44mZ2Z1mVm1m1aNHZ6Srt3POuRRkbZKRNIlgrpJ3mdnLUceTDvVNrdxy/3Ns3NkQdSjOOdcvIqsuC0dKPh+okFRHMFNfAYCZ3QF8HigHfhDOyNpqZtXRRJseD63YxD3LNjJuZDEfu/C4qMNxzrmMi7J32bxu9n8A+EA/hdMv5i/ZAMCiNXE+dmHEwTjnXD/I2uqyXLOybg+rNu1l7PAilm/cRUNza/dvcs65Ac6TTD+5a0ktxQUx/uOyGbS0GUtrdkUdknPOZZwnmX6wv6mV363YzJtPHc8FJ4yhIE8sWrsj6rCccy7jPMn0gwUrNtPQ3Ma8MycxpDCfmZWjWLTGZ3h2zuU+TzL9YP6SDZxwTCkzK0cCcPbUclZt3sOehpaII3POuczyJJNhK+v2sHLTHubNmUTYFZu50yowgyfXeWnGOZfbPMlk2PylGyjKj/GWmRMObju9ciQlBXneLuOcy3meZDKovqmV3y3fxGWnjmdEScHB7YX5MWZPKWPRWi/JOOdymyeZDPr9s5upb27jujMrj9g3d2o5a7btZ+vexggic865/uFJJoPmL9nAcWOHccakUUfsO2dqBQBPemnGOZfDPMlkyKpNe3i27vAG/2Qzxg9nREkB/1zj7TLOudzlSSZD7g4b/N+a1OCfLC8mzj62nEVr45hZP0fnnHP9w5NMBjQ0t/LQ8s286ZRxjBxSeNTjzplWzqbdB9jgQ/8753KUJ5kM+MOzW9jf1Mp1Z3Y9S2eiXeaf/vS/cy5HeZLJgLuWbGD6mGHMmnxkg3+yqaOHMqa0yJ+Xcc7lLE8yabZ6815WbNx91Ab/ZJKYO62CJ9fGaW/3dhnnXO7xJJNmdy/dQGF+jLed0XmDf0dnTy0nXt/My9v2ZTgy55zrf55k0uhAcxu/fWZTtw3+yeZO83YZ51zu8iSTRn94bjP7mlqZN6frBv9kE0aWUFU+hCe9XcY5l4M8yaTR/CUbmDp6KLOrum7w7+jsqRU8tW4nrW3tGYrMOeei4UkmTV58dS/PbOhZg39Hc6eVs6+plec27clQdM45Fw1PMmly95KNFObFePsZE3v93rOPLQd8HDPnXO7xJJMGB5rbePCZOi455RhGDe1Zg3+y8mFFnHBMqY9j5pzLOZ5k0mDhyi3sbexdg39H50ytYFntLhpb2tIYmXPORSuyJCPpZ5K2SVp1lP2SdKukNZKek3RGf8fYU/OXbODYiqGcOaUs5XPMnVZOc2s7z9TuSmNkzjkXrShLMj8HLu5i/yXA9PB1A/DDfoip117euo9ltbtSavBPNmdKGXkx+WyZzrmcElmSMbN/ADu7OOQK4JcWWAyMlDSuf6LruflLNgQN/rN63+CfrLS4gFMnjuCf/ryMcy6HZHObzARgY9J6XbgtazS2tPHgM5u46ORjKEuhwb+juVMreK5uD/saW9IQnXPORS+bk0yPSbpB0jJJy7Zv395v1/3Tqi3sOdDCvDmVaTnfOVPLaWs3lqzvqoDnnHMDRzYnmU1A8l/vieG2I5jZnWZWbWbVo0eP7pfgAOY/tZGq8iEHn3PpqzMmj6IoP+bjmDnnckY2J5kFwLvDXmZnAXvMbEvUQSWs2baPJTU7+9zgn6y4II/qqlE+v4xzLmdE2YV5PvAkcLykOknXS7pR0o3hIQuBdcAa4MfAv0QUaqfmL9lIQZ763ODf0TlTK3jx1X3s2N+U1vM651wU8qO6sJnN62a/AR/up3B6pbGljQeeqeONJx1DxbCitJ77nKlB1dvidXEuO3V8Ws/tnHP9LZury7LWn1e9yu6GFq7rwxP+R3PKhBGUFuV7u4xzLid4kknBXUs2MDmNDf7J8vNinHlsmbfLOOdygieZXlqzbT9L1u/k2tmTiMXS0+Df0dlTK6iNN1C3qyEj53fOuf7SqzYZST0ZnKvdzHanGE/Wu3vJBvJj4so0N/gnmzstKCEtWhvn6uohGbuOc85lWm8b/jeHr67+hc8D0t9YkQUONfiPZXRpehv8kx0/tpTyoYU8uTbO1dXpedDTOeei0Nsk84KZzezqAEnL+xBPVnv4+VfZ1dDSpyH9e0ISZ08t559rdmBmaXsOxznn+ltv22TOTtMxA9L8JRuoLCth7tSKjF9r7rQKtu1rYu32+oxfyznnMqVXScbMGtNxzEC0bvt+Fq/LbIN/ssTzMt7LzDk3kPU4yUh6g6QfSzo9XL8hc2Fln7uXbiQ/Jq6qzlyDf7JJZUOYMLLEp2R2zg1ovSnJvB+4GXinpAuA0zMTUvZpam3j/qfruPDEsYwpLe6Xa0rinKnlLF63k7Z265drOudcuvUmyewzs91m9kngjcDsDMWUdf7y/FZ21jdz3Zn922lu7rQK9hxoYfXmvf16XeecS5feJJk/JhbM7NPAL9MfTnaav2QDE0eVcO60zDf4J/N2GefcQNfjJGNmvwOQVBGu35apoLLJ+h31LFobZ96c/mnwTzZmeDHTxgzjn2t9HDPn3MCUyrAyP0t7FFns7qUbyIuJqzL4hH9X5k4tZ+n6nTS3tkdyfeec64tUksygeTKwubWd+5fVceGJYxgzvH8a/Ds6e2oFB1raWLExZ0fqcc7lsFSSzKDp6vTI6q3E65sz/oR/V84+thwJ78rsnBuQvCTThYWrtjB2eBHnTR8dWQwjhhRw8vgRPOntMs65ASiVJPOZtEeRpdZvr2fGuOHk9XODf0fnTCtn+cZdNDS3RhqHc871Vq+TjJmtykQg2cbMqI3XM7l8aNShMHdqBS1txtKaXVGH4pxzvZLSpGWSqiX9VtIzkp6TtFLSc+kOLko79jdT39xGVXn087lUV42iIE8s8nYZ59wA09uh/hN+QzDEzEogJ/vW1saD0Y8nV0RfkhlSmM/MSaNY5O0yzrkBJtXpl7eb2QIzW29mtYlXWiOLWG08mPq4KguqyyB4+n/V5j3sbmiOOhTnnOuxVJPMFyT9RNI8SW9LvNIaWcRq4/XkxcSEkSVRhwIE45iZweJ1Xppxzg0cqVaXvQ84ASjgUHWZAQ+mI6hsUBNvYPzIYgrzU83D6XXaxJEMKcxj0do4F588LupwnHOuR1JNMrPN7Pi0RpJlauP1WVNVBlCYH2N2VZk/lOmcG1BS/Td9kaQZfbmwpIslvSRpjaRPd7J/kqRHJS0Pe7Bd2pfr9VZNvIHJWdCzLNncaeWs3V7P1r05Ofmocy4HpZpkzgJWhEmi112YJeUBtwOXADOAeZ0krc8B95rZTOBa4Acpxtpruxua2XOgJatKMgDnTA2mGvCh/51zA0Wq1WUX9/G6c4A1ZrYOQNLdwBXA6qRjDBgeLo8ANvfxmj1WE/Ysy4YHMZPNGDecESUFLFoT560zoxkV2jnneiPVksyXgT1JXZf3Al/oxfsnABuT1uvCbcm+SDDVcx2wEPjXFGPttcQzMtnwIGayWEycfWw5i9bGMRs045Q65wawVJPMqWZ2cOx5M9sFzExPSAfNA35uZhOBS4FfSeo0Xkk3SFomadn27dv7fOHaeAMSVJZlV5KBoF1m0+4DB5/jcc65bJZqkolJGpVYkVRG76reNgGVSesTw23JrgfuBTCzJ4FioNP5j83sTjOrNrPq0aP7PmJyTbyeccOLKS7I6/O50u2caYl2GX9exjmX/VJNMv8NPCnpK5K+AiwCvtWL9y8FpkuaIqmQoGF/QYdjNgCvB5B0IkGS6XsxpQdq4w1MyrKqsoRjK4YydngR//TGf+fcAJBSkjGzXwJvA7aGr7eZ2a968f5W4CbgYeAFgl5kz0v6sqTLw8M+AXxQ0rPAfOC91k8NEdn2jEwyScydWsHitXHa271dxjmX3VLtXYaZrebw3mC9ff9Cggb95G2f73D+uameP1X7GlvYsb8563qWJTt7ajkPLt/ES1v3ceK44d2/wTnnIpIdY6ZkkUMDY2ZndRl4u4xzbuDwJNNBbZY+I5NswsgSqsqH+Pwyzrmsl3KSkXRLOgPJFjWJeWSyuCQDQWnmqfU7aW3Lyel8nHM5osdJRtK9Sa/7gA9kMK7IbIg3MLq0iKFFKTdX9Yu5UyvY39TKc5v2RB2Kc84dVW/+ku41s4OJRdIPMxBP5Gri9VndHpNw1rFlACxas4MzJo3q5mjnnItGb6rLvgogKfFA5GfTH070auMNWd0ek1A+rIgTxw33xn/nXFbrcZIxs/Xh4s/C9Z0ZiShCB5rbeHVvI5OzcDiZzpwztZxltbtobGmLOhTnnOtUKg3/SnsUWWLDzrBnWUX2l2QgGMesubWdZ2p3RR2Kc851KpUkk7OPmddk6ejLRzNnSjl5MfkQM865rOUlmSSJIf4nlw2MksywonxOmziCJ17xJOOcy06pJJnPpD2KLFETb2DUkAJGDCmIOpQeu+ikY3i2bg9rtu2POhTnnDtCKknmeUmV3R828NTG6wdEz7JkbztjIvkxcd+yjd0f7Jxz/azXSSYcCXlhtwcOQLXxhgHTHpMwurSIC04YwwPP1NHiT/8757JMqsPKPCNpdlojiVhTaxubdx8YcCUZgGtmV7JjfzN/f3Fb1KE459xhUk0yZxJMWrZW0nOSVkp6Lp2B9be6XQdot+wfs6wzrz1uNGNKi7h3qVeZOeeyS6oDdF2U1iiywMGeZQOwJJOfF+PKWRO547G1vLqnkWNGFEcdknPOAamXZOYAO82sFngX8F2gLG1RRaBmR/bPI9OVq6sraTd44Jm6qENxzrmDUk0y/2Fm+ySdC1wI/BS4I31h9b/aeD2lRfmUDS2MOpSUVFUM5cwpZdy7bKNPy+ycyxqpJpnEYFlvAu40sz8CA/Ovc6gm3sDkiiFIA/dZ02tmV1Ibb+Cp9Tk3rJxzboBKNclskvQj4BpgoaSiPpwrKwzEZ2Q6uuTkcZQW5XOvPzPjnMsSqSaGq4GHgYvMbDdBe8zNaYuqn7W2tVO368CAbY9JKCnM4/LTx7Nw5Rb2HGiJOhznnEstyZhZg5k9aGavhOtbzOwv6Q2t/2ze3Uhruw34kgwEVWZNre0seHZz1KE459zAruJKl0OjLw/8JHPKhBGcOG64PzPjnMsKKScZSbekM5AoHXpGZmBXlwFI4prqiazctIfVm/dGHY5zbpDrcZKRdG/S6z7gA325sKSLJb0kaY2kTx/lmKslrZb0vKS7+nK9rtTEGyguiDGmtChTl+hXb5k5gcL8mHcAcM5Frjclmb1mdnX4ugr4a6oXlZQH3A5cAswA5kma0eGY6QTTCsw1s5OAj6V6ve7UxuupKh86oLsvJxs5pJCLTjqG3y7f5FMzO+ci1Zsk89UO65/tw3XnAGvMbJ2ZNQN3A1d0OOaDwO1mtgvAzDI2+mNNvCEnqsqSXVNdyZ4DLfxl9daoQ3HODWI9TjJmth5AUkW43pcn/iYAyXU5deG2ZMcBx0n6p6TFki7uw/WOqr3d2LCzISca/ZOdM7WciaNKvAOAcy5SqTT8/yztUXQuH5gOnA/MA34saWRnB0q6QdIyScu2b9/eq4u8ureR5tb2nOi+nCwWE1fNquSJNTvYuLMh6nCcc4NUKkkmHQ0Xm4Dk2TUnhtuS1QELzKwlLEW9TJB0jmBmd5pZtZlVjx49uleBHOq+nFvVZQBXVk9EwmfNdM5FJpUkk47RF5cC0yVNkVQIXAss6HDMQwSlmEQV3XHAujRc+zC18eC//MkVuVWSAZgwsoTzpo/mvqfraPNBM51zEYikJGNmrcBNBEPTvADca2bPS/qypMvDwx4G4pJWA48CN5tZvK/X7qgmXk9hXoxjhufmHCzXVFeyZU8jj7/Su2pE55xLh1QmLftMOi5sZguBhR22fT5p2YCPh6+Mqd3RQGVZCXmx3Oi+3NGFM8YwakgB9y7byPnHj4k6HOfcINPrkoyZrZJ0laRSAEmfk/SgpDPSH17m1YTPyOSqovw83jpzIo+s3kp8f1PU4TjnBpl0Tlr2w/SF1T/MjNp4Q871LOvomtmVtLQZv13esW+Fc85l1qCetGz7viYOtLRRVZF7PcuSHX9MKadVjuTeZRsJaiGdc65/DOpJy2oSPctyvCQDQQeAl7fuZ8XG3VGH4pwbRAb1pGW1OfyMTEdvPm0cJQV5Pmimc65fDepJy2rjDeTHxISRJVGHknGlxQW86dRx/P7ZLTQ0t0YdjnNukBhwVVzpVBOvZ+KoEvLzBsdtuGZ2JfubWvnjc1uiDsU5N0gMjr+uR1Ebb2DSIGiPSaiePIpjRw/1KjPnXL8ZtDNjmln4jEzut8ckSOLq6kqW1uxi7fb9UYfjnBsEIpsZM2q7GlrY19g6KHqWJXvbGRPIi8lLM865fhHJzJjZIJdHX+7KmNJiLjhhDA88vYmWtvaow3HO5bioZsaMXKL78mAryUDwzMyO/U38/cWMTTbqnHNAdDNjRq423oAElWW53325o/OPH82Y0iKfNdM5l3HZPDNmRtXGGxg/ooSi/LyoQ+l3+Xkx3j5rIo++tI2texujDsc5l8OimhkzcjXx+pwfs6wrV1dX0m5w/9N1UYfinMthUc2MGbnaeAOTygZfe0zClIqhzJlSxn0+aKZzLoMGZUlmz4EWdtY3D7qeZR1dU11JTbyBp9YP6OY151wWSyXJpGVmzChtGESjL3fl0lPGUVqU7x0AnHMZk9LMmB23STpX0u3pCSnzDj4jM4jbZABKCvN48+njWbhqC3sbW6IOxzmXg7pMMpLyJc2SdEQXLEkzJX1bUg3wFeDFDMWYdolnZCaVDe4kA0GVWWNLOwtWbI46FOdcDuquJHMzMB24BUDScZK+IOlF4DZgAyAze52Z3ZbZUNOnNt7A2OFFDCnMjzqUyJ06cQQnHFPqw8w45zKiuyTzLeB54Nvh+ovABcBlZnZumFjajvbmbFUbbxj07TEJkrhmdiXP1e3hhS17ow7HOZdjukwyZtZmZivNLFFh/zZgC/CopB9Lej0DsLfZYBt9uTtvOX0ChXkx7vEOAM65NOtVw7+ZPWRm1wInAI8CHwPGSPqhpDdmIsB0a2huZdu+Ji/JJBk1tJA3njSWh1Zsoql1wBVMnXNZLNXpl+vN7C4zezMwEVhO2G6T7WoPdl/2kkyya2ZXsruhhb88vzXqUJxzOaTPM2Oa2S4zu9PMXt+b90m6WNJLktZI+nQXx71dkkmq7muscKhnWZWXZA4zd2oFE0aWeAcA51xaRTL9ctgl+nbgEmAGME/SjE6OKwU+CjyVrmvXhCWZSV6SOUwsJq6qnsjjr+xg486GqMNxzuWIlJKMpI938rpe0uk9PMUcYI2ZrTOzZuBu4IpOjvsK8E0gbUMF18brKR9ayPDignSdMmdcVV2JBPf5oJnOuTRJtSRTDdwITAhf/w+4GPixpE/14P0TgOR6mbpw20GSzgAqzeyPKcbYqaD7spdiOjNhZAnnTqvg/mUbaWv3QTOdc32XapKZCJxhZp8ws08As4AxwGuA9/Y1KEkx4DvAJ3p4/A2Slklatn379i6PrY03eHtMF66ZXcnmPY38edWrUYfinMsBqSaZMUBT0noLMNbMDnTYfjSbgMqk9YnhtoRS4GTg/8Jha84CFhyt8T/seFBtZtWjR48+6kUbW9rYvOeAd1/uwhtnHMOMccP53EMreXWPT2jmnOubVJPMb4CnwiFmvgD8E7hL0lBgdQ/evxSYLmmKpELgWmBBYqeZ7TGzCjOrMrMqYDFwuZktSzFeAOp2NWDmA2N2pTA/xm3XzaSxpZ1/u2eFV5s55/qk10lGkoCfAzcAu8PXjWb25fD5mXd0dw4zawVuAh4GXgDuNbPnJX1Z0uW9jamnanaEPct8YMwuTR09jC9dcRJProtzx2Nrow7HOTeA9XqESDMzSQvN7BQg5ZKFmS0EFnbY9vmjHHt+qtdJVuPPyPTYVbOC7szfeeRlzjq2nFmTR0UdknNuAEq1uuwZSbPTGkk/qI03MLw4n5FDvHWMjTUAABV3SURBVPtydyTx1beezPiRxXxk/nL2HPD5ZpxzvZdqkjkTWCxpraTnJK2U9Fw6A8uEmng9VRVDCWr8XHeGFxdw67Uz2bq3kX9/cCVm3j7jnOudVCdUuSitUfST2ngDp1WOjDqMAWXmpFF8/I3H8a0/v8RrllVwzexJUYfknBtAUi3JbADOA95jZrWAAWPTFlUGtLS1s2n3AR/iPwU3vmYqc6eV88UFq1mzbV/U4TjnBpBUk8wPgLOBeeH6PoKxyLLWpl0HaGs3f0YmBbGY+M7Vp1NSmMdNdy2nscWnA3DO9UzKbTJm9mHCMcXMbBdQmLaoMuBQzzIvyaRi7PBi/vuq03jx1X18408vRh2Oc26ASDXJtIQjKRuApNFAe9qiyoBD88h4SSZVrzthDO+fO4WfL6rhr6t93hnnXPdSTTK3Ar8lmBXzq8ATwNfSFlUG1MTrGVKYR8WwrC5wZb1bLjmeGeOGc/P9z/qwM865bqU6M+ZvgE8BXwe2AG8xs/vSGVi6BaMve/flvirKz+O262bS1OrDzjjnupfypGVm9qKZ3W5m3zezF9IZVCbUxOu9PSZNpo4expcuD4ad+eH/rYk6HOdcFutVkpH0TDqO6W9t7cbGnQ3eHpNGV86ayOWnjee7f32Fp2t3Rh2Ocy5L9fZhzBO7ebJfwIg+xJMRW/YcoKXNvCSTRpL4z7eezPKNu/jI/BUs/Oh5jCjx4Xqcc4frbZI5oQfHZN1DFN6zLDMSw85cdceT/PuDK/n+dTO9zcs5d5heJZnw6f4B5+AzMj6PTNolDztz3tIKrp3jw8445w5JueF/IKmNN1CUH2NsaXHUoeSkG18zlXOnVfDF3z/vw8445w4zKJJMzY56JpUNIRbzqpxMCIadOY0hhfk+7Ixz7jBpSTKSxkkqSse5MiHxjIzLnDE+7IxzrhPpKsn8CnhR0n+l6Xxp095u1O70Z2T6gw8745zrKC1JxswuBI4F/jcd50unbfuaaGxpZ3KFl2T6wy2XHM9J433YGedcIG1tMhZ4Pl3nS5daH325XxXl53HbvGDYmY/ds9yHnXFukEtrw7+kKklvTec5+yrxjEyVt8n0m2PDYWcWr9vpw844N8j1OclIukXSI5IWAx8CsqoDQE28noI8MW6Ed1/uTz7sjHMO0lOSaQPeDPwDWG9md6fhnGlTG2+gctQQ8vMGRW/trJEYdmb8yGI+Mn8Few60RB2Scy4CvR0g84eSPiTpXEnDAczsv8ys0cw+BayV9KOMRJqimng9k709JhKJYWe27m3ko3cvZ1+jJxrnBpve/nu/HJgBfBVYL6lW0u8lfU3StWb2CHBLT04k6WJJL0laI+nTnez/uKTVkp6T9DdJk3sZK2bmz8hEbOakUXzpipN4/JUdXHbbEzy7cXfUITnn+lGvkoyZ3Wlm/2pmrzWzcmAu8ANgD/Cm8Jhu/4qEUzffDlxCkLTmSZrR4bDlQLWZnQrcD3yrN7ECxOub2d/U6iWZiL3jzMncc8NZtLYZb//hIn78j3W0e68z5waFPjVUmFmdmf3JzL5pZu/qxVvnAGvMbJ2ZNQN3A1d0OPejZtYQri4GJvY2vkPdl70kE7XqqjIWfuQ8Xn/iGL668AXe/4ul7NjfFHVYzrkMi6o1fAKwMWm9Ltx2NNcDf+rtRWp2JIb495JMNhgxpIA73jmLr7zlZBatjXPp9x5n0ZodUYflnMugrO9yJemdQDXw7S6OuUHSMknLtm/ffnB77c4GYoKJozzJZAtJvOusyfzuw3MZXlLAO376FN9++EVa29qjDs05lwFRJZlNQGXS+sRw22EkXQh8FrjczI5atxK2FVWbWfXo0aMPbq+N1zNhVAmF+VmfSwedE8cNZ8FNc7l6ViW3P7qWa+5cTN2uhu7f6JwbUKL667sUmC5piqRC4FpgQfIBkmYCPyJIMNtSuUhNvMHbY7LYkMJ8vnnlqdw6byYvvbqPS7/3OH9etSXqsJxzaRRJkjGzVuAm4GHgBeBeM3te0pclXR4e9m1gGHCfpBWSFhzldEdV68/IDAiXnzaehR85jykVQ7nx18/wuYdW+pw0zuWIXk2/nE5mthBY2GHb55OWL+zL+Xc3NLO7oYXJZV6SGQgmlQ/hvhvP4b//8hI/+sc6ltXs4vvXzWTamNKoQ3PO9UHONlYkBsb0kszAUZgf4zOXnsjP3zeb7fuauOy2J7hn6QbM/Jka5waqnE0yNYlnZHwemQHn/OPH8KePnsesyaO45YGV/Ov85ez1IWmcG5ByNskkSjKTyrwkMxCNGV7Mr95/JjdfdDx/WvUqb7r1cVb4kDTODTg5nWTGjSimuCAv6lBcimIx8eHXTePe/3c27e1w5Q8X8aPH1vqQNM4NIDmcZLxnWa6YNXkUCz96Hm88aSxf/9OLvPfnPiSNcwNFziYZf0Ymt4woKeD2687ga289hafWxbnke4/z0yfWs6fB22qcy2Y5mWT2N7WyY3+TD/GfYyRx3ZmTWHDTuUwuG8JX/rCaM7/+V26+71lWbNztvdCcy0KRPSeTSYnRl726LDcdf0wp93/oHFZv3stvnqrloeWbuO/pOk4aP5x3njWZy08bz9CinPzRdm7AycmSjD8jMzjMGD+cr771FJ767IX851tOpq3d+MyDKznza3/jPx5axUuv7os6ROcGvZz8d6/mYEnGq8sGg2FF+bzzrMm848xJPLNhN79ZXMs9yzbyq8W1VE8exTvPmszFJx/jPQ2di0BOJpnaHQ1UDCtimFeZDCqSmDV5FLMmj+I/LpvBA8/U8ZunNvCxe1Yw6vcFXF1dybw5k/wBXef6kXKtsbS6utqm3XAbrW3G/R86J+pwXMTa240n18X59eJa/rJ6K23txnnTK3jHmZO58MQx5OflZI2xc70i6Wkzq87EuXPyX/3aeAPnTK2IOgyXBWIxMXdaBXOnVbB1byP3LN3I/CUbuPHXTzN2eBHXzp7EtXMqGTeiJOpQnctJOZdkzGDLnkaqvNHfdTB2eDEfef10/uX8qTz60nZ+81Qtt/79FW77+yu8/sSxvP2MiZw5pYxRQwujDtW5nJFzSaapNZjGd7LXu7ujyM+L8YYZY3nDjLFsiDcwf+kG7l26kUdWbwVg+phhzJ5SxpyqMmZPKWPCSC/lOJeqnEsyza3BZFdeknE9Mal8CLdcfAL/duFxPFu3myXrd7K0Zie/X7GZu57aAMCEkSXMrhp1MPFMGzMMSRFH7tzAkHtJpi0syfhkZa4XCvNjzK4qY3ZVGQBt7caLr+5l6fqdLKnZyRNr4jy0YjMAo4YUUF11qKRz0vjhFHgHAuc6lXNJpqm1nbFDChgxpCDqUNwAlhcTJ40fwUnjR/DeuVMwM2riDQeTztKanQer14YU5jFz0khmh4ln5qRRlBT6MznOQQ4mmebWdn8I06WdJKZUDGVKxVCunl0JwNa9jSyt2Rkmnl1872+vYAb5MXHyhBHMmVLG9DHDOHb0UKrKh1I2tNCr2dygk3NJpqm13dtjXL8YO7yYy04dz2Wnjgdgz4EWnq7dyZL1u1has5P//ed6WtoOPYdWWpzPsRVDqaoIkk4i+VRVDGVEiZe8XW7KuSTT0uYlGReNESUFXHDCWC44YSwQ/CzW7TrA+h37Wb+jgZod9azfUc+yml0seHYzyc9Blw8tpCosKU0Jk9CUiqFUVQxhSGHO/Zq6QSQnf3q9JOOyQUFe7GDS6KixpY0NOxtYv6P+YPJZv6Oex1/Zzv1P1x127NjhRQfPM3HUEEaXFgWvYcHX8qGFPnKBy1o5mWS8JOOyXXFBHseNLeW4saVH7KtvaqUmXk/NjoZDpaB4PQ8/v5Wd9c1HHC9B2ZBCRpcWUREmnkQSqigtZPSw4oPbRpYUEIt5u5DrPzmZZLwk4wayoUX5B3u2dXSguY0d+5vYvr+J7fsOvXYk1vc3UVNTz/Z9TQcfTE6WFxMVww4lpIphQeIZXlJAaXE+w4s7LudTWlxAaVG+JyeXkpxLMjGJMh8WxOWoksI8KsuGUFnW9T9SZsb+ptZDiWh/Ezv2HZ6cduxv5sUt+9jb2EJDc1uX55OCKRWGF4cJqKSA4UdJSkOL8hlSmEdJQT4lhXnhct7B5eL8PE9Yg0hkSUbSxcD3gDzgJ2b2jQ77i4BfArOAOHCNmdV0d97C/Jh3E3WDnqSgBFJcwLGjh3V7fEtbO/sbW9nb2MLeA8HXfUnLextb2XugJdweLG/a3cgLB/axr7GFfU2t9GZA9+KCGEMK8w9PPgXB18OX8ykuyKMoP3bwVZgfoyg/j8L8GIV5MYoKEl/zDl9PPi4/Rp4ntkhEkmQk5QG3A28A6oClkhaY2eqkw64HdpnZNEnXAt8Erunu3EX53gDqXG8V5MUYNbQw5cFB29uN/c2t7GtsZX9jKwda2mhobuVAc1u43HbYcmO4/9By8NpZ30zdruRjW2lsObLaLxX5MYUJKkZBXuIl8pOXYzq4Lz9PSduTjw/WC/Nj5MeC9xfmiVgseH9MIi+x3GFb4tXZtryYyOtsW4ftMQUxHLYt6Zz54bZs+Wc7qpLMHGCNma0DkHQ3cAWQnGSuAL4YLt8PfF+SrJsJcAo9yTjX72IxBdVlxel/3qe93Whua6e5rZ2mlsTXtmBbaztNrYmvbQfXD20Lvnbc39zaTkt7O61tRmt7Oy1tRktbsN7S1k5Dcyut7Za0PemY9uBr4vjW9uyck0visMQTC5djCpoVlLScSVElmQnAxqT1OuDMox1jZq2S9gDlwI6uTlzkXTmdyymxmCiOBVVoFEcdzZHMgmTUbkZbe5B02hNfw22JV2fb2jquJ7a1Hb6v3YzWo2zr6bUB2s3CVxB7ezs8lcH7kxMN/5JuAG4AmFA5OeJonHODiSQK87OjaipV38rguaP6t38TUJm0PjHc1ukxkvKBEQQdAI5gZneaWbWZVR8zxmfEdM65bBFVklkKTJc0RVIhcC2woMMxC4D3hMtXAn/vrj3GOedcdomkuixsY7kJeJigC/PPzOx5SV8GlpnZAuCnwK8krQF2EiQi55xzA0hkbTJmthBY2GHb55OWG4Gr+jsu55xz6eNdsZxzzmWMJxnnnHMZ40nGOedcxniScc45lzGeZJxzzmWMcu3RE0nbgdqo4+hnFXQz3M4g4/fjSH5PDuf343DHm9mRM+ilQU4MK5PMzEZHHUN/k7TMzKqjjiNb+P04kt+Tw/n9OJykZZk6t1eXOeecyxhPMs455zLGk0xuuDPqALKM348j+T05nN+Pw2XsfuRcw79zzrns4SUZ55xzGeNJJotI+pmkbZJWJW0rk/SIpFfCr6PC7ZJ0q6Q1kp6TdEbSe94THv+KpPckbZ8laWX4nluVLZOAd0JSpaRHJa2W9Lykj4bbB+X9AJBULGmJpGfDe/KlcPsUSU+Fn+OecPoMJBWF62vC/VVJ5/pMuP0lSRclbb843LZG0qf7+zOmQlKepOWS/hCuD/b7URP+XK9I9BqL9PfGzPyVJS/gNcAZwKqkbd8CPh0ufxr4Zrh8KfAnQMBZwFPh9jJgXfh1VLg8Kty3JDxW4Xsvifozd3EvxgFnhMulwMvAjMF6P8J4BQwLlwsIZs09C7gXuDbcfgfwoXD5X4A7wuVrgXvC5RnAs0ARMAVYSzDlRl64fCxQGB4zI+rP3YP78nHgLuAP4fpgvx81QEWHbZH93nhJJouY2T8I5s5JdgXwi3D5F8Bbkrb/0gKLgZGSxgEXAY+Y2U4z2wU8Alwc7htuZost+En5ZdK5so6ZbTGzZ8LlfcALwAQG6f0ACD/b/nC1IHwZcAFwf7i94z1J3Kv7gdeH/3VeAdxtZk1mth5YA8wJX2vMbJ2ZNQN3h8dmLUkTgTcBPwnXxSC+H12I7PfGk0z2G2tmW8LlV4Gx4fIEYGPScXXhtq6213WyPeuF1RozCf5zH9T3I6waWgFsI/jFXwvsNrPW8JDkz3Hws4f79wDl9P5eZbP/AT4FtIfr5Qzu+wHBPx5/kfS0pBvCbZH93uTcE/+5zMxM0qDqDihpGPAA8DEz25tc/TsY74eZtQGnSxoJ/BY4IeKQIiPpMmCbmT0t6fyo48ki55rZJkljgEckvZi8s79/b7wkk/22hkVUwq/bwu2bgMqk4yaG27raPrGT7VlLUgFBgvmNmT0Ybh609yOZme0GHgXOJqjiSPzDmPw5Dn72cP8IIE7v71W2mgtcLqmGoCrrAuB7DN77AYCZbQq/biP4R2QOEf7eeJLJfguARM+O9wC/S9r+7rB3yFnAnrA4/DDwRkmjwh4kbwQeDvftlXRWWA/97qRzZZ0wxp8CL5jZd5J2Dcr7ASBpdFiCQVIJ8AaCtqpHgSvDwzrek8S9uhL4e1iPvgC4NuxtNQWYTtCYuxSYHvbOKiRoHF+Q+U+WGjP7jJlNNLMqglj/bmbvYJDeDwBJQyWVJpYJft5XEeXvTZS9IPx1RK+Q+cAWoIWgrvN6gjrjvwGvAH8FysJjBdxOUCe/EqhOOs/7CRov1wDvS9peHf7ArQW+T/gwbja+gHMJ6pafA1aEr0sH6/0I4z0VWB7ek1XA58PtxxL8UVwD3AcUhduLw/U14f5jk8712fBzv0RS76DwHr8c7vts1J+5F/fmfA71Lhu09yP87M+Gr+cTMUf5e+NP/DvnnMsYry5zzjmXMZ5knHPOZYwnGeeccxnjScY551zGeJJxzjmXMZ5knIuYpHxJ/x4+i+FcTvEk41z0ZgDrgdOiDsS5dPMk41z0XgSmEjxw6lxO8YcxnXPOZYyXZJxLkaRTJNVK+lAPjq2SdCAcpj95+1skmaQTkraVhLMaNkuqyETszvUXTzLOpcjMVhIMmvjuHr5lrZmd3mHbPOCJ8GvivAfC4zanJVDnIuRJxrm+2QaclMobw7lyziUYCPXadAblXLbwJONc33wDKJI0OYX3XgH82cxeBuKSZqU3NOei50nGuRRJugQYCvyR1Eoz8wgm2yL8Oq+LY50bkHz6ZedSIKkY+CZwOfA+4GRgYS/eX0Ywk+Mp4VS4eYBJutm8y6fLIV6ScS41nwN+aWY1BJM9nZzYIelvkiZ08/4rgV+Z2WQzqzKzSoIHMs/LVMDORcGTjHO9JOl4gqmP/yfcdDDJSIoB04Cd3ZxmHsH868kewKvMXI7x6jLnesnMXgLO7LB+Rrg6A3jAzA50c47XdbLt1nTG6Vw28JKMc2lkZqvM7OOd7GoDRnR8GLMziYcxgQKgPd0xOteffFgZ55xzGeMlGeeccxnjScY551zGeJJxzjmXMZ5knHPOZYwnGeeccxnjScY551zGeJJxzjmXMZ5knHPOZYwnGeeccxnz/wFK/8LF9QZA9gAAAABJRU5ErkJggg==\n",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ }
+ }
+ ],
+ "source": [
+ "lam = np.linspace(nu.max().to(u.AA, equivalencies=u.spectral()),\n",
+ " nu.min().to(u.AA, equivalencies=u.spectral()), 1000)\n",
+ "bb_lam = BlackBody(bb.temperature, \n",
+ " scale=1.0 * u.erg / (u.cm ** 2 * u.AA * u.s * u.sr))\n",
+ "bb5000K_lam = bb_lam(lam)\n",
+ "plt.plot(lam, bb5000K_lam)\n",
+ "plt.xlim([1.0e3, 5.0e4])\n",
+ "plt.xlabel(r'$\\lambda$, [{0:latex_inline}]'.format(lam.unit))\n",
+ "plt.ylabel(r'$I_{\\lambda}$, ' + '[{0:latex_inline}]'.format(bb5000K_lam.unit))\n",
+ "plt.title('Planck function in wavelength')\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 38
+ },
+ "id": "A4VLHFtHXNbN",
+ "outputId": "01a83a99-4944-4d30-968a-0375a0877477"
+ },
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/latex": "$1.146776 \\times 10^{10} \\; \\mathrm{\\frac{erg}{s\\,sr\\,cm^{2}}}$",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "execution_count": 5
+ }
+ ],
+ "source": [
+ "np.trapz(x=lam, y=bb5000K_lam).to('erg s-1 cm-2 sr-1')"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "-BC40k7BXNbN"
+ },
+ "source": [
+ "Observe que isso está dentro de alguns por cento da resposta que obtivemos no espaço de frequência, apesar de nossa amostragem ruim em pequenos comprimentos de onda!\n",
+ "\n",
+ "Muitas funções `astropy` usam unidades e quantidades diretamente. À medida que você ganha confiança ao trabalhar com eles, considere incorporá-los ao seu fluxo de trabalho regular. Leia mais [aqui](http://docs.astropy.org/en/stable/units/) sobre como usar unidades."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "rlwI6oPGXNbO"
+ },
+ "source": [
+ "###Como simular observações reais\n",
+ "\n",
+ "Desde outono de 2017, o `astropy` não suporta explicitamente a construção de observações sintéticas de modelos como curvas de corpo negro. A [biblioteca synphot](https://synphot.readthedocs.io/en/latest/) permite isso. Você pode usar `synphot` para realizar tarefas como transformar espectros em magnitudes visuais por convolução com uma curva de filtro."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "88l25iyEXNbO"
+ },
+ "source": [
+ "## A função de massa inicial estelar (IMF)\n",
+ "\n",
+ "A função de massa inicial estelar nos diz quantas estrelas de cada massa são formadas. Em particular, as estrelas de baixa massa são muito mais abundantes do que as estrelas de alta massa. Vamos explorar mais a funcionalidade do `astropy` usando esta ideia.\n",
+ "\n",
+ "As pessoas geralmente pensam no FMI como uma função de densidade de probabilidade de lei de potência. Em outras palavras, se você contar as estrelas que nasceram recentemente de uma nuvem de gás, sua distribuição de massas seguirá o FMI. Vamos escrever uma pequena classe para nos ajudar a acompanhar isso:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {
+ "id": "NToG17WYXNbO"
+ },
+ "outputs": [],
+ "source": [
+ "class PowerLawPDF(object):\n",
+ " def __init__(self, gamma, B=1.):\n",
+ " self.gamma = gamma\n",
+ " self.B = B\n",
+ " def __call__(self, x):\n",
+ " return x**self.gamma / self.B"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "gqNFgJqNXNbP"
+ },
+ "source": [
+ "### O método `__call__` \n",
+ "\n",
+ "Ao definir o método `__call__`, estamos dizendo ao interpretador Python que uma instância da classe pode ser chamada como uma função. Quando chamada, uma instância desta classe recebe um único argumento, `x`, mas usa outros atributos da instância, como `gamma` e `B`.\n",
+ "\n",
+ "### Mais sobre as aulas\n",
+ "\n",
+ "As classes são estruturas de dados mais avançadas, que podem ajudá-lo a acompanhar a funcionalidade em seu código que funciona em conjunto. Você pode aprender mais sobre aulas neste [tutorial](https://www.codecademy.com/ja/courses/learn-python/lessons/introduction-to-classes/exercises/why-use-classes)."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "SPGk79FOXNbP"
+ },
+ "source": [
+ "## Integração usando quadratura gaussiana\n",
+ "\n",
+ "Nesta seção, exploraremos um método de integração numérica que não requer já ter sua grade de amostragem configurada. `scipy.integrate.quad` com referência [aqui](https://docs.scipy.org/doc/scipy-0.19.1/reference/generated/scipy.integrate.quad.html) recebe uma função e um valor menor e limite superior, e nossa classe `PowerLawPDF` cuida disso muito bem.\n",
+ "\n",
+ "Agora podemos usar nossa nova classe para normalizar nosso IMF dados os limites de massa. Isso equivale a normalizar uma função de densidade de probabilidade. Usaremos a quadratura gaussiana (`quad`) para encontrar a integral. `quad` retorna o valor numérico da integral e sua incerteza. Nós só nos importamos com o valor numérico, então vamos empacotar a incerteza em `_` (uma variável de espaço reservado). Nós imediatamente jogamos a integral em nosso objeto IMF e a usamos para normalizar!\n",
+ "\n",
+ "Para ler mais sobre *empacotamento e desempacotamento generalizado* em Python, veja a proposta original, [PEP 448](https://www.python.org/dev/peps/pep-0448/), que foi aceita em 2015."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 287
+ },
+ "id": "8_PJmxbYXNbP",
+ "outputId": "35681e7d-0473-4b2d-dc24-5deba60a4037"
+ },
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEOCAYAAACXX1DeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd3hUddrG8e+ThN6kBKUKSBMJKkSkJha6BUUs2F0VQbHAFvXVXcuuq65rQBAUC6KuiqygIhaKhUiTJtJ7EbAAotJrnvePDO/mZZNxQjKZmcz9ua65kvObmXNuf4Z55pwz8xxzd0RERPKSEOkAIiIS3VQoREQkKBUKEREJSoVCRESCUqEQEZGgVChERCSopEgHCIdq1ap5vXr1Ih1DRCSmzJ8/f7u7Jx87XiwLRb169Zg3b16kY4iIxBQz25jbuA49iYhIUCoUIiISlAqFiIgEpUIhIiJBqVCIiEhQKhQiIhKUCkUO23YdYOuu/ZGOISISVaK+UJjZJWb2opm9bWZdwrmthz9YSueMTP49bxO6ToeISLaIFAozG2VmW81syTHj3cxspZmtMbP7ANz9PXe/FegHXBnOXIM6N6bxieX54zuLuH7UHDbt2BvOzYmIxIRI7VGMBrrlHDCzRGA40B1oBvQxs2Y5HvJg4P6wOSW5PG/3bcujPU9jwcaf6Tokk9Ez1pOVpb0LEYlfESkU7p4J7DhmuDWwxt3XuftBYAzQ07I9CXzs7gvyWqeZ9TWzeWY2b9u2bcedLSHBuL5tPSYNTCO1XhUe/mAZl4+cxZqtu457nSIisSyazlHUAjblWN4cGLsT6AT0NrN+eT3Z3V9w91R3T01O/q+eVvlWu3JZXr3pLJ6+/HTWbN1Nj2emM/zzNRw6klXgdYuIxJKobwro7kOBoZHYtplxWavapDVO5uEJS3lq0komLvqep3q3oHmtSpGIJCJS5KJpj2ILUCfHcu3AWMQlVyjF8Gta8vy1rdi++wA9h8/giY9XsP/QkUhHExEJu2gqFHOBRmZW38xKAlcBEyKc6f/p1vwkpg5Mp3fL2jw/bS09nvmSOeuPPdUiIlK8ROrjsW8Bs4AmZrbZzG5298PAAGASsBwY6+5LI5EvmEplS/Bk7xb86+azOXgkiytGzuLP7y1h94HDkY4mIhIWVhy/WJaamupFceGivQcP889Jq3hl5npqVCzNY71SOLdJ9bBvV0QkHMxsvrunHjseTYeeYk7Zkkn85aJmjOvfjnKlkrjplbkMenshP+85GOloIiKFRoWiELSsW5mJd3XgrvMaMuGb7+iUMY2Ji75TGxARKRZUKApJqaREBnVpwgd3dqDmCWUY8ObX3Pb6fH7cqSaDIhLbVCgK2ak1KvLu7e24r3tTpq3aRqeMabw991vtXYhIzFKhCIOkxAT6pZ/CJ/ekcWqNitw7bjHXvawmgyISm1Qowqh+tXKMubUNf7ukOQs3/UKXwZm8PH09R9RkUERiiApFmCUkGNe2OZnJA9No06AKf524jN7Pz2T1j2oyKCKxQYWiiNQ8oQyjbjyLIVeewYbte7hg6HSGfrqag4fVZFBEopsKRREyMy45sxZTBqXTtflJZExZxcXPTmfR5l8iHU1EJE8qFBFQrXwphvU5kxevT+XnvQe5ZPgMHv9oOfsOqsmgiEQfFYoI6tzsRKYMSufKs+owMnMd3Z/JZPa6nyIdS0Tk/1GhiLCKpUvweK8WvHnL2WQ5XPXCbB54dzG79h+KdDQREUCFImq0a1iNSfekcWvH+rw151u6DM7ksxU/RjqWiIgKRTQpUzKRBy5oxvjb21OxdAl+N3oed4/5mp92H4h0NBGJYyoUUeiMOifwwZ0duKdTIz5a/D2dB2cy4Rs1GRSRyFChiFIlkxK4p1NjJt7ZkTpVynLXW19z62vz+eFXNRkUkaKlQhHlmpxUgfH92/FAj1OZvmYbnTOm8dYcNRkUkaKjQhEDEhOMW9MaMOmeNJrXqsT94xdz9YtfsfGnPZGOJiJxQIUihpxctRxv3no2j/dKYcmWX+k6JJOXvlynJoMiElYqFDHGzOjTui5TBqXToWE1/vbhcno9N5OVP6jJoIiEhwpFjDqpUmlevD6VoX3OZNOOvVw47EuGTF2lJoMiUuhUKGKYmXHx6TWZOiidC1JqMGTqai4aNp2Fm9RkUEQKjwpFMVClXEmGXHUmo25MZef+Q/QaMYO/TVymJoMiUihUKIqR85qeyOSBafRpXZeXpq+n65BMZq7ZHulYIhLjVCiKmQqlS/DYpSmM6duGBIOrX/qK+8cvYqeaDIrIcVKhKKbaNKjKx3encVt6A96eu4nOGdOYskxNBkUk/1QoirEyJRO5v/upvHdHeyqXLcmtr81jwJsL2K4mgyKSDyoUcaBF7ROYMKADgzo3ZvLSH+mcMY33vt6iNiAiEhIVijhRMimBu85vxId3deDkquW45+2F3PzqPL77ZV+ko4lIlFOhiDONTqzAuP7t+POFzZi19ie6DM7kX7M3kqU2ICKSBxWKOJSYYNzcoT6T7knj9DqVePC9JfR5cTbrt6vJoIj8NxWKOFa3aln+dfPZPHlZCsu+30m3IZmMnLaWw0fUBkRE/iPqC4WZlTOzV83sRTO7JtJ5ihsz48qz6jJ1UDppjZN5/OMV9HpuJsu/3xnpaCISJSJSKMxslJltNbMlx4x3M7OVZrbGzO4LDPcC3nH3W4GLizxsnDixYmleuK4Vz159Jt/9so+Lhk0nY/JKDhxWGxCReBepPYrRQLecA2aWCAwHugPNgD5m1gyoDWwKPEyvWmFkZlzYoiZTBqZz8ek1GfrZGi4cOp0F3/4c6WgiEkERKRTungnsOGa4NbDG3de5+0FgDNAT2Ex2sYAgec2sr5nNM7N527ZtC0fsuFG5XEkyrjyDV246iz0HDnPZczN59INl7D14ONLRRCQCoukcRS3+s+cA2QWiFjAeuMzMngM+yOvJ7v6Cu6e6e2pycnJ4k8aJc5tUZ/KgdK5rczKjZqyny+BMpq9Wk0GReBNNhSJX7r7H3W9y9/7u/kak88Sb8qWSeLRnc8be1pYSiQlc+/JX/Omdb/h1r5oMisSLaCoUW4A6OZZrB8YkCrSuX4WP7+5I/3NOYdyCLXQaPI1PlvwQ6VgiUgSiqVDMBRqZWX0zKwlcBUyIcCbJoXSJRO7t1pT372hPcvlS9PvXfG5/Yz5bd+2PdDQRCaNIfTz2LWAW0MTMNpvZze5+GBgATAKWA2PdfWkk8klwzWtV4v0B7flj1yZMXbaVzhmZjJu/WU0GRYopK47/uFNTU33evHmRjhEX1mzdzb3jFjF/48+kN07m771SqHVCmUjHEpHjYGbz3T312PFoOvQkMahh9fKMva0tD1/UjLkbdtAlYxqvzdqgJoMixYgKhRRYYoJxY/vsJoMtT67MX95fypUvzGLttt2RjiYiheA3C4WZpRRFEIl9daqU5bXfteap3i1Y+cMuuj/zJSO+WMMhNRkUiWmh7FGMMLM5Zna7mVUKeyKJaWbG5al1mDoonfOaVOcfn6zkkuEzWLLl10hHE5Hj9JuFwt07AteQ/R2H+Wb2ppl1DnsyiWnVK5bm+eta8dw1Lflx5wF6Dp/BU5NWsP+Q2nWJxJqQzlG4+2rgQeBeIB0YamYrzKxXOMNJ7OueUoOpg9LodWYthn++lh5Dv2TehmPbfIlINAvlHEULMxtM9ncbzgMucvdTA78PDnM+KQZOKFuSpy4/ndd+15oDh7K4fOQsHp6wlD0H1GRQJBaEskcxDFgAnO7ud7j7AgB3/47svQyRkKQ1TmbywDRuaFuPV2dtoMvgTKatUqdfkWgXSqF4191fd/d9RwfM7G4Ad389bMmkWCpXKomHLz6Nf9/WltIlErhh1Bx+P/Ybftl7MNLRRCQPoRSK63MZu7GQc0icSa1XhQ/v6siAcxvy/sItdMrI5KPF30c6lojkItiFgPqY2QdAfTObkOP2Of990SGRfCtdIpE/dG3C+wPac1KlUtz+xgL6vT6frTvVZFAkmiQFuW8m8D1QDXg6x/guYFE4Q0l8Oa1mJd67vT0vfrmewVNXMTNjOw9e2IzLW9XGzCIdTyTuqSmgRJV127KbDM7d8DMdG1Xj75emUKdK2UjHEokL+W4KaGbTAz93mdnOHLddZrYznGElfjVILs/bfdvyaM/TWLDxZ7oOyeSVGevVZFAkgvIsFO7eIfCzgrtXzHGr4O4Viy6ixJuEBOP6tvWYNDCNs+pV4ZEPlnH5yFms2bor0tFE4lIoX7g7xcxKBX4/x8zuMrMTwh9N4l3tymUZfdNZZFxxOmu37abHM9N59rPVajIoUsRC+XjsOOCImTUEXiC759ObYU0lEmBm9GpZmykD0+l82on8c/IqLn5WTQZFilIohSIrcJnSS4Fh7v5HoEZ4Y4n8f8kVSjH86paMvK4VP+3ObjL4xMdqMihSFEIpFIfMrA9wAzAxMFYifJFE8tb1tJOYMjCd3i1r8/y0tfR45kvmrNfXekTCKZRCcRPQFnjM3debWX1ArTskYiqVLcGTvVvwxi1ncygriytGzuLP7y1h1/5DkY4mUizpexQS0/YePMw/J63ilZnrqVGxNI/1SuHcJtUjHUskJuX7exQ5ntjezKaY2SozW2dm681sXXhiiuRP2ZJJ/OWiZozr345ypZK46ZW5DHp7IT/vUZNBkcLym3sUZrYCGAjMB/7vzKG7/xTeaMdPexTx6cDhIwz/fC0jPl9DpTIleKTnaVyQUkNtQERCdNx7FMCv7v6xu29195+O3sKQUaRASiUlMqhzYz64swO1KpdhwJtf0/f1+fyoJoMiBRJKofjczJ4ys7Zm1vLoLezJRI7TqTUqMr5/O/6nR1MyV22jU8Y03p77LcXxfJxIUQjl0NPnuQy7u58XnkgFp0NPctSG7Xu4d9wivlq/g/YNq/L4pS2oW1VNBkVyk9ehJ33qSYq9rCznrbnf8vhHKziS5fyhaxNubFePxASduxDJqSCfejrRzF42s48Dy83M7OZwhBQJh4QE45qzT2bywDTaNKjCXycuo/fzM1n9o5oMioQilHMUo4FJQM3A8irgnnAFEgmXmieUYdSNZ/HMVWewYfseLhg6naGfrubgYTUZFAkmlEJRzd3HAlkAgb5ParAjMcnM6HlGLaYOSqdr85PImLKKi5+dzjebfol0NJGoFUqh2GNmVQEHMLM2gFp3SkyrWr4Uw/qcyYvXp/Lz3oNcOmIGj3+0nH0H9R5I5FjBrpl91CBgAnCKmc0AkoHeYU0lUkQ6NzuRsxtU4fGPljMycx2Tlv7A471a0PaUqpGOJhI1QvrUk5klAU0AA1a6e5F2XzOzS4ALgIrAy+4+Odjj9aknOR4z12znvvGL+XbHXq4+uy73dW9KxdJqlCzxI98fjzWzXsFW6O7jQ9zwKOBCYKu7N88x3g14BkgEXnL3J0JYV2Xgn+4e9FNXKhRyvPYdPELGlJW8PH091SuU5rFLm3P+qSdGOpZIkTieQvFK4NfqQDvgs8DyucBMd78wxA2nAbuB144WCjNLJPvTU52BzcBcoA/ZRePxY1bxO3ffGnje08Ab7r4g2DZVKKSgFm76hXvfWcTKH3fR84ya/OXCZlQtXyrSsUTC6ri/cGdmk4Eb3P37wHINYLS7d83HxusBE3MUirbAw0fXYWb3A7j7sUXi6PMNeAKY4u5T83hMX6AvQN26dVtt3Lgx1HgiuTp4OIvnvljLs5+vpkLpEjx0UTMuPr2mmgxKsVWQpoB1jhaJgB+BugXMUwvYlGN5c2AsL3cCnYDeZtYvtwe4+wvunuruqcnJyQWMJwIlkxK4u1MjJt7ZkTpVynL3mIXc+to8vv91X6SjiRSpUArFp2Y2ycxuNLMbgQ+BXN/Vh4u7D3X3Vu7ez92fL8ptizQ5qQLj+7fjwQtOZfqa7XTJyOTNr74lK6v4tb8Ryc1vFgp3HwA8D5weuL3g7ncWcLtbgDo5lmsHxkSiUmKCcUvHBky6J43mtSrxP+8u5uqXZrPxpz2RjiYSdkXSFDCXcxRJZJ/MPp/sAjEXuNrdlxbG9nQyW8LJ3RkzdxN//3A5h7Ky+EOXJtzUvr6aDErMK8g5ioJu+C1gFtDEzDab2c2BNiADyO4htRwYW1hFQiTczIw+resyZVA6HRpW428fLqfXczNZ+YOaDErxpDbjIgXg7nyw6HsembCUnfsPcfs5Dbnj3IaUTAr7ezCRQleQNuMXmZn+6kVyYWZcfHpNpgxK54KUGjzz6WouHPYlC9VkUIqRUArAlcBqM/uHmTUNdyCRWFSlXEmGXHUmo25MZdf+w/QaMYO/TVymJoNSLITyqadrgTOBtcBoM5tlZn3NrELY04nEmPOansjkgWn0aV2Xl6avp+uQTGau2R7pWCIFEtIhJXffCbwDjAFqAJcCC8ysoB+TFSl2KpQuwWOXpjCmbxsSDK5+6SvuG7eIX/cVaS9NkUITyjmKnmb2LvAFUAJo7e7dyf5Oxe/DG08kdrVpUJVP7knjtvQGjJ23ic4Z05i89IdIxxLJt1D2KHoBg909xd2fOtqgz933Arp2tkgQpUskcn/3U3nvjvZUKVeSvq/PZ8CbC9i++0Cko4mELJRC8YO7Z+YcMLMnAdz907CkEilmWtQ+gQkDOvD7zo2ZvPRHOmVM492vN1McP54uxU8ohaJzLmPdCzuISHFXMimBO89vxId3daB+tXIMfPsbbho9ly2/qMmgRLc8C4WZ9TezxUBTM1uU47YeWFR0EUWKl0YnVuCdfu34y4XN+GrdDrpkTOP12RvVZFCiVrALF1UCKpN9IaH7cty1y913FEG246ZvZkus2LRjL/ePX8z0NdtpXb8KT17WgvrVykU6lsSp4/lmtrv7BuAOYFeOG2ZWJRwhReJNnSplef3m1vzjshYs/34n3YZk8vy0tRw+khXpaCL/J1iheDPwcz4wL/Bzfo5lESkEZsYVZ9Xh00HppDdO5omPV3DpiJks+25npKOJAGoKKBJV3J2Pl/zAX95fwi97D9H/nFMYcF5DSiUlRjqaxIG8Dj0lBXlCy2ArdPcFhRFMRP7DzOiRUoO2Dary1w+XMeyzNXy85AeevKwFrU6uHOl4EqeCncz+PMjz3N3PC0+kgtMehRQXX6zcygPvLuG7X/dxY7t6/KFLE8qVyvP9nUiB5LVHoUNPIlFu94HD/OOTFbw2ayO1K5fh8V4pdGyUHOlYUgzlu1CY2Xnu/pmZ9crtfncfX8gZC40KhRRHc9bv4L5xi1i3fQ9XpNbmgR7NqFS2RKRjSTGS73MUQDrwGXBRLvc5ELWFQqQ4al2/Ch/d3ZGhn65mZOY6Pl+5jb/2bE635idFOpoUczr0JBKDlmz5lT+9s4hl3++kR8pJPHJxc5IrlIp0LIlxBbkUalUzG2pmC8xsvpk9Y2ZVwxNTRELRvFYl3h/Qnj92bcLU5VvplDGNcfPVZFDCI5SmgGOAbcBlQO/A72+HM5SI/LYSiQnccW5DPrqrIw2rl+f3//6GG16Zy+af90Y6mhQzv3noycyWuHvzY8YWu3tKWJMVgA49SbzJynJen72RJz9ZAcC93ZpyXZuTSUiwCCeTWHLch56AyWZ2lZklBG5XAJMKP6KIHK+EBOOGdvWYdE8arU6uzEMTlnLFyFms3bY70tGkGAj28dhdZH+6yYBywNEuZQnAbnevWCQJj4P2KCSeuTvjFmzhrxOXse/QEe4+vxF90xpQIjGU94USz/K9R+HuFdy9YuBngrsnBW4J0VwkROKdmdG7VW2mDErj/KbVeWrSSi4ZPoMlW36NdDSJUSG9xTCzymbW2szSjt7CHUxECqZ6hdI8d20rnrumJT/uPEDP4TP4xycr2H/oSKSjSYwJ5eOxtwCZZJ+XeCTw8+HwxhKRwtI9pQafDkrn0jNrMeKLtfQY+iVzN0T1tcckyoSyR3E3cBaw0d3PBc4EfglrKhEpVJXKluCfl5/Oa79rzYFDWVwxchYPvb+E3QcORzqaxIBQCsV+d98PYGal3H0F0CS8sUQkHNIaJzN5YBo3tK3Ha7M30nVwJtNWbYt0LIlyoRSKzWZ2AvAeMMXM3gc2hjeWiIRLuVJJPHzxabzTry2lSyRww6g5/H7sN/yy92Cko0mUylevJzNLByoBn7h71P5V6eOxIqHZf+gIz362huenreWEsiV4tGdzeqTUiHQsiZCCfOEOM2tpZncBLYDN0VwkRCR0pUsk8oeuTXh/QHtOqlSa299YQL/X57N15/5IR5MoEsqnnv4CvApUBaoBr5jZg+EOdkyGcmY2z8wuLMrtisSL02pW4r3b23Nvt6Z8tjK7yeDYeZvUZFCA0PYorgHOcveH3P0hoA1wXSgrN7NRZrbVzJYcM97NzFaa2Rozuy+EVd0LjA1lmyJyfJISE+h/zil8cndHmp5UkT+9s4jrR81h0w41GYx3oRSK74DSOZZLAVtCXP9ooFvOATNLBIYD3YFmQB8za2ZmKWY28ZhbdTPrDCwDtoa4TREpgAbJ5RnTtw1/7XkaCzb+TNchmbwyYz1HsrR3Ea/yvMKdmQ0ju9fTr8BSM5sSWO4MzAll5e6eaWb1jhluDaxx93WB7YwBerr748B/HVoys3PI7jXVDNhnZh+5e9axjxORwpOQYFzXth7nnXoi/zN+MY98sIyJi77nyctSaFi9QqTjSRELdinUox8bmg+8m2P8iwJusxawKcfyZuDsvB7s7g8AmNmNwPa8ioSZ9QX6AtStW7eAEUUEoNYJZRh901m8+/UWHp24jB7PTOeu8xtyW/opajIYR/IsFO7+6tHfzawk0DiwuNLdD4U7WC55Rv/G/S8AL0D2x2OLIpNIPDAzerWsTcdGyTz8wVL+OXkVHy7+gad6t6B5rUqRjidFIJRPPZ0DrCb7vMIIYFUBmwJuAerkWK5N6Oc8RCRCkiuUYvjVLRl5XSt+2p3dZPCJj9VkMB6Esu/4NNDF3dPdPQ3oCgwuwDbnAo3MrH5gT+UqYEIB1iciRajraScxZVA6vVvW5vlpa+nxzJfMWa8mg8VZKIWihLuvPLrg7quAEqGs3MzeAmYBTcxss5nd7O6HgQFkd6FdDox196X5jy4ikVKpTAme7N2CN245m0NZ2U0G//zeEnbtL/Kj0lIEQrlm9ivAEeBfgaFrgER3/12Ysx03tfAQKTp7Dx7m6cmrGDVjPTUqluaxXimc26R6pGPJcShIC49+ZH+P4a7AbRnQv3DjiUisKlsyiT9f2Ixx/dtRrlQSN70yl4FvL2THHnX6KS6C7lEEvhy31N2bFl2kgtMehUhkHDh8hOGfr2XE52uoVKYEj/Q8jQtSamBmkY4mITiuPQp3PwKsNDN9MUFEflOppEQGdW7MB3d2oFblMgx482v6vj6fH9VkMKaFcuipMtnfzP7UzCYcvYU7mIjErlNrVGR8/3b8T4+mZK7aRqeMaYyZ862aDMaoUE5mp+c27u7TwpKoEOjQk0j02LB9D/eNX8TsdTtod0pVHu+VwslVy0U6luQir0NPeRYKMytN9onshsBi4OXAR1ujngqFSHTJynLGzN3E3z9azuGsLP7QpQk3ta9PYoLOXUST4zlH8SqQSnaR6E72F+9ERPItIcG4+uy6TBmURrtTqvG3D5dz2XMzWfXjrkhHkxAEKxTN3P1adx8J9AY6FlEmESmmalQqw8s3pPLMVWew8ac9XDD0S4Z+upqDh9UQOpoFKxT/9xXLWDnkJCLRz8zoeUYtpg5Kp1vzGmRMWcXFz07nm02/RDqa5CFYoTjdzHYGbruAFkd/N7OdRRVQRIqnquVLMazPmbx0fSq/7D3EpSNm8PePlrPvoJoMRptgbcYTizKIiMSnTs1OpHWDKjz+0QpeyFzHpKU/8ESvFrQ9pWqko0mArjwiIhFXsXQJHu+Vwpu3Zl/DrM+Ls7l//GJ2qslgVFChEJGo0e6Uanxydxq3dqzP23O/pUtGJp8u/zHSseKeCoWIRJUyJRN54IJmjL+9PZXKlODmV+dx11tf89PuA5GOFrdUKEQkKp1R5wQ+uLMDAzs15uMl39N5cCbvL9yiNiARoEIhIlGrZFICd3dqxMQ7O1KnSlnuHrOQW16dx/e/7ot0tLiiQiEiUa/JSRUY378dD15wKjPWbqdzRiZvfLWRrCztXRQFFQoRiQmJCcYtHRsw+Z50WtSuxAPvLuHql2azYfueSEcr9lQoRCSm1K1aljduOZsneqWwdMtOug7J5IXMtRw+ojYg4aJCISIxx8y4qnVdpgxKp2OjZP7+0Qoue24mK35Q04hwUKEQkZh1UqXSvHh9K4b1OZPNP+/jwqHTyZiyigOH1QakMKlQiEhMMzMuOr0mUwalc9HpNRn66WouGjadr7/9OdLRig0VChEpFqqUK8ngK89g1I2p7Np/mF7PzeSvE5ex96CaXxeUCoWIFCvnNT2RyQPTuLp1XV6evp5uQ75k5prtkY4V01QoRKTYqVC6BI9dmsLbfduQmGBc/dJX3DduEb/uU5PB46FCISLF1tkNqvLx3R25Lb0BY+dtonPGNCYv/SHSsWKOCoWIFGulSyRyf/dTee+O9lQpV5K+r89nwJsL2K4mgyFToRCRuNCidnaTwT90aczkpT/SKWMa7369WU0GQ6BCISJxo0RiAgPOa8RHd3egQbVyDHz7G24aPZctv6jJYDAqFCISdxpWr8C/+7XjoYua8dW6HXTJmMbrs9VkMC8qFCISlxITjJva12fywDRanlyZP7+3hKtemM26bbsjHS3qqFCISFyrU6Usr/2uNf/o3YIVP+yk2zNf8twXajKYU9QXCjNLMLPHzGyYmd0Q6TwiUvyYGVek1mHqoHTObZLMk5+s4JIRM1j2nZoMQpgLhZmNMrOtZrbkmPFuZrbSzNaY2X2/sZqeQG3gELA5XFlFRKpXLM3z17ZixDUt+eHX/Vz87HSenrwy7psMhnuPYjTQLeeAmSUCw4HuQDOgj5k1M7MUM5t4zK060ASY6e6DgP5hzisicc7M6JFSgykD07n4jJoM+2wNFwydzvyN8dtkMKyFwt0zgR3HDLcG1rj7Onc/CIwBerr7Yne/8JjbVrRtRLIAAAsuSURBVLL3Io7+H4rvsi4iRaZyuZJkXHEGo286i30Hj9D7+Zk88sFS9hyIvyaDkThHUQvYlGN5c2AsL+OBrmY2DMjM60Fm1tfM5pnZvG3bthVOUhGJe+c0qc6kgWlc3+ZkXpmxga5DMvlydXy9xkT9yWx33+vuN7v7ne4+PMjjXnD3VHdPTU5OLsqIIlLMlS+VxCM9m/Pvfm0pmZTAdS/P4Y///oZf98ZHk8FIFIotQJ0cy7UDYyIiUe2selX46K6O3H7OKYz/egudBk/jkyXFv8lgJArFXKCRmdU3s5LAVcCECOQQEcm30iUS+VO3prx/R3uSy5ei37/mc/sb89m6a3+ko4VNuD8e+xYwC2hiZpvN7GZ3PwwMACYBy4Gx7r40nDlERApb81qVeH9Ae/7UrQlTl2+lc0Ym78wvnk0GrTj+R6Wmpvq8efMiHUNE4sSarbu5b9wi5m38mbTGyfz90ubUrlw20rHyzczmu3vqseNRfzJbRCTaNaxenrG3teXRnqcxb8MOugzO5NWZG4pNk0EVChGRQpCQYFzfth6TB6aRWq8KD01YyhUjZ7G2GDQZVKEQESlEtSuX5dWbzuLpy09n9dbddH/mS4Z/voZDMdxkUIVCRKSQmRmXtarN1EHpdDq1Ok9NWsklw2ewZMuvkY52XFQoRETCJLlCKUZc04rnr23J1l0H6Dl8Bv/4ZAX7D8VWNyIVChGRMOvWvAZTB6bT68xajPhiLT2GfsncDce2wYteKhQiIkWgUtkSPHX56bx+c2sOHs7iipGzeOj9JeyOgSaDKhQiIkWoY6NkJt2Txo3t6vHa7I10HZzJtFXR3WRQhUJEpIiVK5XEQxedxjv92lGmZCI3jJrDoLEL+WXvwUhHy5UKhYhIhLQ6uTIf3tWBO89ryISF39EpYxofLf4+0rH+iwqFiEgElUpK5PddmjBhQAdqVCrD7W8s4LbX57F1Z/Q0GVShEBGJAs1qVuTd29txf/emfLFyG50ypjF27qaoaDKoQiEiEiWSEhO4Lf0UPrknjaY1KvKncYu47uU5bNqxN6K5VChERKJM/WrlGHNrG/52SXMWbvqFLoMzGTV9PUci1GRQhUJEJAolJBjXtjmZyQPTaNOgCo9OXMblz89k9Y+7ij5LkW9RRERCVvOEMoy68SwGX3k667bv4YKh0xn66WoOHi66JoMqFCIiUc7MuPTM7CaDnU87kYwpq7j42eks2vxLkWxfhUJEJEZUK1+K4Ve3ZOR1rdix5yCXDJ/B4x8tD3uTQRUKEZEY0/W0k5gyKJ0rUuswMnMd3YZkMnvdT2HbngqFiEgMqlSmBE9c1oI3bjmbI+5c9cJsHnh3Mbv2Hyr0balQiIjEsPYNqzHpnjRu7lCft+Z8y+IwXBwpqdDXKCIiRapsyST+fGEzbmxXjzpVyhb6+rVHISJSTISjSIAKhYiI/AYVChERCUqFQkREglKhEBGRoFQoREQkKBUKEREJSoVCRESCsmi4zF5hM7NtwMbAYiXg2K8qHjuWc7kasD1M0XLLUljPCfa4vO4LZW5yG4vm+Qr1eYU1X7mNa76C3xfv8xXs/kjP18nunvxfo+5erG/AC781lnMZmFeUWQrrOcEel9d9ocxNrM1XqM8rrPn6rfmJ5/nK6754n69g90frfMXDoacPQhjL7THhcDzbCfU5wR6X132hzE1uY9E8X6E+r7DmK7dxzVfw++J9voLdH5XzVSwPPRWEmc1z99RI54gVmq/80Xzlj+Yrf8I1X/GwR5FfL0Q6QIzRfOWP5it/NF/5E5b50h6FiIgEpT0KEREJSoVCRESCUqEQEZGgVCjywcwuMbMXzextM+sS6TzRzswamNnLZvZOpLNEKzMrZ2avBv6urol0nminv6n8KazXrLgpFGY2ysy2mtmSY8a7mdlKM1tjZvcFW4e7v+futwL9gCvDmTfSCmm+1rn7zeFNGn3yOXe9gHcCf1cXF3nYKJCf+YrXv6mc8jlfhfKaFTeFAhgNdMs5YGaJwHCgO9AM6GNmzcwsxcwmHnOrnuOpDwaeV5yNpvDmK96MJsS5A2oDmwIPO1KEGaPJaEKfLzm++SrQa1bS8T4x1rh7ppnVO2a4NbDG3dcBmNkYoKe7Pw5ceOw6zMyAJ4CP3X1BeBNHVmHMV7zKz9wBm8kuFguJrzdu/yef87WsaNNFn/zMl5ktpxBes+LyDzOHWvzn3Rxk/6OtFeTxdwKdgN5m1i+cwaJUvubLzKqa2fPAmWZ2f7jDRbm85m48cJmZPUfRta6IBbnOl/6m8pTX31ehvGbFzR5FYXD3ocDQSOeIFe7+E9nHRiUP7r4HuCnSOWKF/qbyp7Bes+J9j2ILUCfHcu3AmORO83X8NHf5o/nKn7DOV7wXirlAIzOrb2YlgauACRHOFM00X8dPc5c/mq/8Cet8xU2hMLO3gFlAEzPbbGY3u/thYAAwCVgOjHX3pZHMGS00X8dPc5c/mq/8icR8qSmgiIgEFTd7FCIicnxUKEREJCgVChERCUqFQkREglKhEBGRoFQoREQkKBUKEREJSoVCRESCUqGQYsHMHjCzpWa2yMwWmtnZZnaCmd0e4vN35/Z7LDCzema2z8wW5hi7zczczM7JMXZHYKxzkHWVCczfQTOrFuboEiNUKCTmmVlbsq+H0dLdW5DdVnkTcAIQUqEowLbNzKLh39Fadz8jx3IK8A3QFMDMygK3ANuARXmtxN33BdbzXRizSoyJhj9wkYKqAWx39wMA7r7d3b8j+4ItpwTeIT8FYGbXmtmcwNjIwJXBcmVm75nZ/MCeSt8c4/UCl5x8DVhCjq6dgftWmNloM1tlZm+YWSczm2Fmq82sdbD1W/Y1tD80s2/MbImZXZnbWAhz0gIYQ6BQAHcB/way3P3HkGZV5Ch31023mL4B5cm+QtwqYASQHhivByzJ8bhTyb44UInA8gjg+sDvu3M8bnfgZ5XAzzJkF4SqOdabBbTJJUs94DDZ7+gTgPnAKMDIvkLbezke+1/rBy4DXszxmEq5jeWyzSXHjG0DGgCfkL1n9TVwDjA1xDndAFSL9P9b3aLjpj0KiXnuvhtoBfQl+wXybTO7MZeHnh943NzA8fzzyX4xzctdZvYNMJvsvYZGOe7b6O6z83jeendf7O5ZwFLgU3d3YDHZL+rB1r8Y6GxmT5pZR3f/NY+xPJlZHeAnz74sZnXgj8AwoHFgXUcf18TMRpjZ02ZWM9g6Jb7pCndSLLj7EeAL4AszWwzcEFjOyYBX3f03L6EZOAncCWjr7nvN7AugdI6H7Any9AM5fs/KsZxF4N9cXut391Vm1hLoAfzNzD5190dzGwuy/RT+UxB2Ad3IvqbyEGBBYPvVgPuBPwFlgX+a2Y3ufjDIeiVOaY9CYl7gnXHOd/tnABvJfpGskGP8U7KvHVw98LwqZnZyHqutBPwceBFvCrQp5Ni5rj/wzn6vu/8LeApomdvYb6y7Bf8pFE8BAwKFNGcBSQOeAS4FTgTGAacV1n+cFC/ao5DioDwwzMxOIPv8wBqgr7v/FDiJvAT42N3/aGYPApMDn1Q6BNxBdlE51idAPzNbDqwk+/BQYcpr/SnAU2aWFcjXP4+xYFLIfuHH3SfmGG9G9qEwgNVAF3d/GsDMrgCmFfQ/SoonXbhIJMaZWT1gors3z+fz+gOdyT4k9y93H5fjvg1AqrtvL7ykEqtUKERiXODk9UyyT2Cf8VuP/411lSH7MpvJQIq77yiEiBLjVChERCQoncwWEZGgVChERCQoFQoREQlKhUJERIJSoRARkaBUKEREJCgVChERCUqFQkREgvpfRxS+q0XN8uQAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ }
+ }
+ ],
+ "source": [
+ "salpeter = PowerLawPDF(gamma=-2.35)\n",
+ "salpeter.B, _ = integrate.quad(salpeter, a=0.01, b=100.)\n",
+ "\n",
+ "m_grid = np.logspace(-2., 2., 100)\n",
+ "plt.loglog(m_grid, salpeter(m_grid))\n",
+ "plt.xlabel(r'Stellar mass [$M_{\\odot}$]')\n",
+ "plt.ylabel('Probability density')\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "RotHN5xIXNbP"
+ },
+ "source": [
+ "### Quantas estrelas M existem a mais do que estrelas O?\n",
+ "\n",
+ "Vamos comparar o número de estrelas anãs M (massa inferior a 60% solar) criadas pelo FMI, com o número de estrelas O (massa superior a 15 vezes solar)."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "OCmyGKTSXNbQ",
+ "outputId": "7c90dc41-8192-471f-b9fd-f8ae017acc27"
+ },
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "20936.017868337385\n"
+ ]
+ }
+ ],
+ "source": [
+ "n_m, _ = integrate.quad(salpeter, a=.01, b=.6)\n",
+ "n_o, _ = integrate.quad(salpeter, a=15., b=100.)\n",
+ "print(n_m / n_o)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "_yOghSWOXNbQ"
+ },
+ "source": [
+ "Existem quase 21.000 estrelas de baixa massa nascidas do que estrelas de alta massa!"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "rdzV9nGgXNbQ"
+ },
+ "source": [
+ "### Onde está toda a massa?\n",
+ "\n",
+ "Agora vamos calcular as massas totais relativas para todas as estrelas O e todas as estrelas M nascidas. Para fazer isso, pondere o Salpeter IMF em massa (ou seja, adicione um fator extra de massa à integral). Para fazer isso, definimos uma nova função que toma a velha lei de potência IMF como um de seus argumentos. Como esse argumento não é alterado em toda a integral, ele é passado para a tupla `args` dentro de `quad`. É importante que haja apenas *um* argumento que mude sobre a integral, e que seja o *primeiro* argumento que a função que está sendo integrada aceita.\n",
+ "\n",
+ "Matematicamente, a integral para as estrelas M é\n",
+ "\n",
+ "$$ m^M = \\int_{.01 \\, M_{\\odot}}^{.6 \\, M_{\\odot}} m \\, {\\rm IMF}(m) \\, dm $$\n",
+ "\n",
+ "e equivale a ponderar a função densidade de probabilidade (o FMI) em massa. De maneira mais geral, você encontra o valor de alguma propriedade $\\rho$ que depende de $m$ calculando\n",
+ "\n",
+ "$$ \\rho(m)^M = \\int_{.01 \\, M_{\\odot}}^{.6 \\, M_{\\odot}} \\rho(m) \\, {\\rm IMF}(m) \\, dm $$"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 9,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "QgrkLAWxXNbQ",
+ "outputId": "40a8267a-0e6c-4edc-886c-17f2240b17e0"
+ },
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "20.29197629920483"
+ ]
+ },
+ "metadata": {},
+ "execution_count": 9
+ }
+ ],
+ "source": [
+ "def IMF_m(m, imf):\n",
+ " return imf(m) * m\n",
+ "\n",
+ "m_m, _ = integrate.quad(IMF_m, a=.01, b=.6, args=(salpeter, ))\n",
+ "m_o, _ = integrate.quad(IMF_m, a=15., b=100., args=(salpeter, ))\n",
+ "\n",
+ "m_m / m_o"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "AeEYuH2WXNbR"
+ },
+ "source": [
+ "\n",
+ "\n",
+ "Assim, cerca de 20 vezes mais massa está ligada em estrelas M do que em estrelas O."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "36L_Y2BPXNbR"
+ },
+ "source": [
+ "### Extras\n",
+ "\n",
+ "* Agora compare a luminosidade total de todas as estrelas O com a luminosidade total de todas as estrelas M. Isso requer uma relação massa-luminosidade, como esta que você usará como $\\rho(m)$:\n",
+ "\n",
+ "$$\n",
+ " \\frac{L}{L_{\\odot}} (M) =\n",
+ " \\begin{cases} \n",
+ " \\hfill .23 \\left( \\frac{M}{M_{\\odot}} \\right)^{2.3} \\hfill , \\hfill & .1 < \\frac{M}{M_{\\odot}} < .43 \\\\\n",
+ " \\hfill \\left( \\frac{M}{M_{\\odot}} \\right)^{4} \\hfill , \\hfill & .43 < \\frac{M}{M_{\\odot}} < 2 \\\\\n",
+ " \\hfill 1.5 \\left( \\frac{M}{M_{\\odot}} \\right)^{3.5} \\hfill , \\hfill & 2 < \\frac{M}{M_{\\odot}} < 20 \\\\\n",
+ " \\hfill 3200 \\left( \\frac{M}{M_{\\odot}} \\right) \\hfill , \\hfill & 20 < \\frac{M}{M_{\\odot}} < 100 \\\\\n",
+ " \\end{cases},\n",
+ "$$\n",
+ "\n",
+ "* Pense em quais estrelas estão produzindo a maior parte da luz e quais estrelas têm a maior parte da massa. Como isso pode resultar em dificuldade em inferir massas estelares a partir da luz que elas produzem? Se estiver interessado em saber mais, consulte este [artigo de revisão](https://ned.ipac.caltech.edu/level5/Sept14/Courteau/Courteau_contents.html)."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "vjo7iJ5IXNbR"
+ },
+ "source": [
+ "## Problema desafiador\n",
+ "\n",
+ "* No momento, não estamos preocupados com os limites da lei de potência, mas o FMI deve cair para zero de probabilidade em massas abaixo de 0,01 massas solares e acima de 100 massas solares. Modifique `PowerLawPDF` de uma forma que permita entradas `float` e `numpy.ndarray`.\n",
+ "* Modifique a classe `PowerLawPDF` para usar explicitamente as construções `units` do `astropy`.\n",
+ "* Deduza uma relação entre a taxa de formação de estrelas recente e a luminosidade de $H\\alpha$. Em outras palavras, encontre um valor de $C$ para a função\n",
+ "\n",
+ "$${\\rm SFR \\, [\\frac{M_{\\odot}}{yr}]} = {\\rm C \\, L_{H\\alpha} \\, [\\frac{erg}{s}]} \\, .$$\n",
+ "\n",
+ "* Como isso depende da inclinação e dos pontos finais do FMI?\n",
+ "* Dê uma olhada no Apêndice B de [Hunter & Elmegreen 2004, AJ, 128, 2170](http://adsabs.harvard.edu/cgi-bin/bib_query?arXiv:astro-ph/0408229)\n",
+ "* Que efeito a alteração do índice da lei de potência ou do limite de massa superior do FMI tem sobre o valor de $C$?\n",
+ "* Preveja o efeito sobre o valor de $C$ de usar uma forma diferente do FMI, como Kroupa ou Chabrier (ambos são mais leves na extremidade de baixa massa).\n",
+ "* Se você ainda não está cansado de IMFs, tente definir uma nova classe que implemente uma IMF de lei de potência quebrada (Kroupa) ou log-parabola (Chabrier). Faça os mesmos cálculos acima."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "id": "Sbov-qwiXNbR"
+ },
+ "outputs": [],
+ "source": [
+ ""
+ ]
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3",
+ "language": "python",
+ "name": "python3"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython"
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python"
+ },
+ "colab": {
+ "name": "units-and-integration.ipynb",
+ "provenance": []
+ }
},
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "import numpy as np\n",
- "from scipy import integrate\n",
- "from astropy.modeling.models import BlackBody\n",
- "from astropy import units as u, constants as c\n",
- "import matplotlib.pyplot as plt\n",
- "\n",
- "%matplotlib inline"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "## The Planck function\n",
- "\n",
- "The Planck function describes how a black-body radiates energy. We will explore how to find bolometric luminosity using the Planck function in both frequency and wavelength space."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Let's say we have a black-body at 5000 Kelvin. We can find out the total intensity (bolometric) from this object, by integrating the Planck function. The simplest way to do this is by approximating the integral using the trapezoid rule. Let's do this first using the frequency definition of the Planck function.\n",
- "\n",
- "We'll define a photon frequency grid, and evaluate the Planck function at those frequencies. Those will be used to numerically integrate using the trapezoidal rule. By multiplying a `numpy` array by an `astropy` unit, we get a `Quantity`, which is effectively a combination of one or more numbers and a unit."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "
\n",
- "**Note on printing units**: \n",
- "\n",
- "Quantities and units can be printed to strings using the [Format String Syntax](https://docs.python.org/3/library/string.html#format-string-syntax). This demonstration uses the `latex_inline` format that is built in to the `astropy.units` package. To see additional ways to format quantities, see the [Getting Started](http://docs.astropy.org/en/stable/units/#getting-started) section of the astropy.units documentation pages.\n",
- "
"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "bb = BlackBody(5000. * u.Kelvin)\n",
- "\n",
- "nu = np.linspace(1., 3000., 1000) * u.THz\n",
- "bb5000K_nu = bb(nu)\n",
- "plt.plot(nu, bb5000K_nu)\n",
- "plt.xlabel(r'$\\nu$, [{0:latex_inline}]'.format(nu.unit))\n",
- "plt.ylabel(r'$I_{\\nu}$, ' + '[{0:latex_inline}]'.format(bb5000K_nu.unit))\n",
- "plt.title('Planck function in frequency')\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### Using $LaTeX$ for axis labels\n",
- "\n",
- "Here, we've used $LaTeX$ markup to add nice-looking axis labels. To do that, we enclose $LaTeX$ markup text in dollar signs, within a string `r'\\$ ... \\$'`. The `r` before the open-quote denotes that the string is \"raw,\" and backslashes are treated literally. This is the suggested format for axis label text that includes markup.\n",
- "\n",
- "Now we numerically integrate using the trapezoid rule."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "np.trapz(x=nu, y=bb5000K_nu).to('erg s-1 cm-2 sr-1')"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Now we can do something similar, but for a wavelength grid. We want to integrate over an equivalent wavelength range to the frequency range we did earlier. We can transform the maximum frequency into the corresponding (minimum) wavelength by using the `.to()` method, with the addition of an *equivalency*."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "lam = np.linspace(nu.max().to(u.AA, equivalencies=u.spectral()),\n",
- " nu.min().to(u.AA, equivalencies=u.spectral()), 1000)\n",
- "bb_lam = BlackBody(bb.temperature, \n",
- " scale=1.0 * u.erg / (u.cm ** 2 * u.AA * u.s * u.sr))\n",
- "bb5000K_lam = bb_lam(lam)\n",
- "plt.plot(lam, bb5000K_lam)\n",
- "plt.xlim([1.0e3, 5.0e4])\n",
- "plt.xlabel(r'$\\lambda$, [{0:latex_inline}]'.format(lam.unit))\n",
- "plt.ylabel(r'$I_{\\lambda}$, ' + '[{0:latex_inline}]'.format(bb5000K_lam.unit))\n",
- "plt.title('Planck function in wavelength')\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "np.trapz(x=lam, y=bb5000K_lam).to('erg s-1 cm-2 sr-1')"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Notice this is within a couple percent of the answer we got in frequency space, despite our bad sampling at small wavelengths!\n",
- "\n",
- "Many `astropy` functions use units and quantities directly. As you gain confidence working with them, consider incorporating them into your regular workflow. Read more [here](http://docs.astropy.org/en/stable/units/) about how to use units."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### How to simulate actual observations\n",
- "\n",
- "As of Fall 2017, `astropy` does not explicitly support constructing synthetic observations of models like black-body curves. The [synphot library](https://synphot.readthedocs.io/en/latest/) does allow this. You can use `synphot` to perform tasks like turning spectra into visual magnitudes by convolving with a filter curve."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "## The stellar initial mass function (IMF)\n",
- "\n",
- "The stellar initial mass function tells us how many of each mass of stars are formed. In particular, low-mass stars are much more abundant than high-mass stars are. Let's explore more of the functionality of `astropy` using this concept.\n",
- "\n",
- "People generally think of the IMF as a power-law probability density function. In other words, if you count the stars that have been born recently from a cloud of gas, their distribution of masses will follow the IMF. Let's write a little class to help us keep track of that:"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "class PowerLawPDF(object):\n",
- " def __init__(self, gamma, B=1.):\n",
- " self.gamma = gamma\n",
- " self.B = B\n",
- " def __call__(self, x):\n",
- " return x**self.gamma / self.B"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### The `__call__` method\n",
- "\n",
- "By defining the method `__call__`, we are telling the Python interpreter that an instance of the class can be called like a function. When called, an instance of this class, takes a single argument, `x`, but it uses other attributes of the instance, like `gamma` and `B`.\n",
- "\n",
- "### More about classes\n",
- "\n",
- "Classes are more advanced data structures, which can help you keep track of functionality within your code that all works together. You can learn more about classes in [this tutorial](https://www.codecademy.com/ja/courses/learn-python/lessons/introduction-to-classes/exercises/why-use-classes)."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "## Integrating using Gaussian quadrature\n",
- "\n",
- "In this section, we'll explore a method of numerical integration that does not require having your sampling grid set-up already. `scipy.integrate.quad` with reference [here](https://docs.scipy.org/doc/scipy-0.19.1/reference/generated/scipy.integrate.quad.html) takes a function and both a lower and upper bound, and our `PowerLawPDF` class takes care of this just fine.\n",
- "\n",
- "Now we can use our new class to normalize our IMF given the mass bounds. This amounts to normalizing a probability density function. We'll use Gaussian quadrature (`quad`) to find the integral. `quad` returns the numerical value of the integral and its uncertainty. We only care about the numerical value, so we'll pack the uncertainty into `_` (a placeholder variable). We immediately throw the integral into our IMF object and use it for normalizing!\n",
- "\n",
- "To read more about *generalized packing and unpacking* in Python, look at the original proposal, [PEP 448](https://www.python.org/dev/peps/pep-0448/), which was accepted in 2015."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "salpeter = PowerLawPDF(gamma=-2.35)\n",
- "salpeter.B, _ = integrate.quad(salpeter, a=0.01, b=100.)\n",
- "\n",
- "m_grid = np.logspace(-2., 2., 100)\n",
- "plt.loglog(m_grid, salpeter(m_grid))\n",
- "plt.xlabel(r'Stellar mass [$M_{\\odot}$]')\n",
- "plt.ylabel('Probability density')\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### How many more M stars are there than O stars?\n",
- "\n",
- "Let's compare the number of M dwarf stars (mass less than 60% solar) created by the IMF, to the number of O stars (mass more than 15 times solar)."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "n_m, _ = integrate.quad(salpeter, a=.01, b=.6)\n",
- "n_o, _ = integrate.quad(salpeter, a=15., b=100.)\n",
- "print(n_m / n_o)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "There are almost 21000 as many low-mass stars born as there are high-mass stars!"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### Where is all the mass?\n",
- "\n",
- "Now let's compute the relative total masses for all O stars and all M stars born. To do this, weight the Salpeter IMF by mass (i.e., add an extra factor of mass to the integral). To do this, we define a new function that takes the old power-law IMF as one of its arguments. Since this argument is unchanged throughout the integral, it is passed into the tuple `args` within `quad`. It's important that there is only *one* argument that changes over the integral, and that it is the *first* argument that the function being integrated accepts.\n",
- "\n",
- "Mathematically, the integral for the M stars is\n",
- "\n",
- "$$ m^M = \\int_{.01 \\, M_{\\odot}}^{.6 \\, M_{\\odot}} m \\, {\\rm IMF}(m) \\, dm $$\n",
- "\n",
- "and it amounts to weighting the probability density function (the IMF) by mass. More generally, you find the value of some property $\\rho$ that depends on $m$ by calculating\n",
- "\n",
- "$$ \\rho(m)^M = \\int_{.01 \\, M_{\\odot}}^{.6 \\, M_{\\odot}} \\rho(m) \\, {\\rm IMF}(m) \\, dm $$"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "def IMF_m(m, imf):\n",
- " return imf(m) * m\n",
- "\n",
- "m_m, _ = integrate.quad(IMF_m, a=.01, b=.6, args=(salpeter, ))\n",
- "m_o, _ = integrate.quad(IMF_m, a=15., b=100., args=(salpeter, ))\n",
- "\n",
- "m_m / m_o"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "So about 20 times as much mass is tied up in M stars as in O stars."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### Extras\n",
- "\n",
- "* Now compare the total luminosity from all O stars to total luminosity from all M stars. This requires a mass-luminosity relation, like this one which you will use as $\\rho(m)$:\n",
- "\n",
- "$$\n",
- " \\frac{L}{L_{\\odot}} (M) =\n",
- " \\begin{cases} \n",
- " \\hfill .23 \\left( \\frac{M}{M_{\\odot}} \\right)^{2.3} \\hfill , \\hfill & .1 < \\frac{M}{M_{\\odot}} < .43 \\\\\n",
- " \\hfill \\left( \\frac{M}{M_{\\odot}} \\right)^{4} \\hfill , \\hfill & .43 < \\frac{M}{M_{\\odot}} < 2 \\\\\n",
- " \\hfill 1.5 \\left( \\frac{M}{M_{\\odot}} \\right)^{3.5} \\hfill , \\hfill & 2 < \\frac{M}{M_{\\odot}} < 20 \\\\\n",
- " \\hfill 3200 \\left( \\frac{M}{M_{\\odot}} \\right) \\hfill , \\hfill & 20 < \\frac{M}{M_{\\odot}} < 100 \\\\\n",
- " \\end{cases},\n",
- "$$\n",
- "\n",
- "* Think about which stars are producing most of the light, and which stars have most of the mass. How might this result in difficulty inferring stellar masses from the light they produce? If you're interested in learning more, see [this review article](https://ned.ipac.caltech.edu/level5/Sept14/Courteau/Courteau_contents.html)."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "## Challenge problems\n",
- "\n",
- "* Right now, we aren't worried about the bounds of the power law, but the IMF should drop off to zero probability at masses below .01 solar masses and above 100 solar masses. Modify `PowerLawPDF` in a way that allows both `float` and `numpy.ndarray` inputs.\n",
- "* Modify the `PowerLawPDF` class to explicitly use `astropy`'s `units` constructs.\n",
- "* Derive a relationship between recent star-formation rate and $H\\alpha$ luminosity. In other words, find a value of $C$ for the function\n",
- "\n",
- "$${\\rm SFR \\, [\\frac{M_{\\odot}}{yr}]} = {\\rm C \\, L_{H\\alpha} \\, [\\frac{erg}{s}]} \\, .$$\n",
- "\n",
- "* How does this depend on the slope and endpoints of the IMF?\n",
- "* Take a look at Appendix B of [Hunter & Elmegreen 2004, AJ, 128, 2170](http://adsabs.harvard.edu/cgi-bin/bib_query?arXiv:astro-ph/0408229)\n",
- "* What effect does changing the power-law index or upper mass limit of the IMF have on the value of $C$?\n",
- "* Predict the effect on the value of $C$ of using a different form of the IMF, like Kroupa or Chabrier (both are lighter on the low-mass end).\n",
- "* If you're not tired of IMFs yet, try defining a new class that implements a broken-power-law (Kroupa) or log-parabola (Chabrier) IMF. Perform the same calculations as above."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": []
- }
- ],
- "metadata": {
- "kernelspec": {
- "display_name": "Python 3",
- "language": "python",
- "name": "python3"
- },
- "language_info": {
- "codemirror_mode": {
- "name": "ipython"
- },
- "file_extension": ".py",
- "mimetype": "text/x-python",
- "name": "python",
- "nbconvert_exporter": "python"
- }
- },
- "nbformat": 4,
- "nbformat_minor": 2
-}
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
\ No newline at end of file