aula9_listas_modulos.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "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": "3.7.0"
    },
    "colab": {
      "name": "aula9_listas_modulos.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/phmouras/Python_Exercises/blob/main/aula9_listas_modulos.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nPetphOMiusH"
      },
      "source": [
        "# Listas (continuação), Módulos e Arquivos\n",
        "\n",
        "Até aqui vimos que listas são coleções ordenadas de sequencias  de valores. Vimos:\n",
        "* como criar listas;\n",
        "* como acessar elementos de uma lista;\n",
        "* comprimento de uma lista;\n",
        "* listas são **mutáveis**; \n",
        "* ***fatias*** de lista;\n",
        "* métodos de lista.\n",
        "\n",
        "\n",
        "Hoje veremos:\n",
        "\n",
        "* referências a listas;\n",
        "* **clonagem** de listas;\n",
        "* listas como parâmetros de funções;\n",
        "* funções **puras** e **modificadoras**."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DZHxpAOhmdr7"
      },
      "source": [
        "## Objetos e referências\n",
        "\n",
        "Se executarmos os seguintes comandos:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OaJDEFFmiusH"
      },
      "source": [
        "a = \"banana\"\n",
        "b = \"banana\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_6e_gr_ziusK"
      },
      "source": [
        "Sabemos que ***a*** e ***b*** se referirão a uma ***string*** mas não sabemos se eles apontam para o mesmo objeto de Python.\n",
        "\n",
        "O interpretador pode organizar sua memória de duas maneiras:\n",
        "\n",
        "* referir ***a*** e ***b*** a dois objetos distintos com os mesmos valores\n",
        "* referir ***a*** e ***b*** ao mesmo objeto\n",
        "\n",
        "![refstring](https://github.com/malbouis/Python_intro/blob/master/aulas_2019/pics/ref_string.png?raw=1)\n",
        "\n",
        "Podemos testar se duas variáveis se referem ao mesmo objeto:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kXF72v-_iusL"
      },
      "source": [
        "a is b"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nOoAV9qQiusN"
      },
      "source": [
        "Nesse exemplo abaixo, ***a*** e ***b*** têm o mesmo valor mas não se referem ao mesmo objeto.\n",
        "\n",
        "![refstring](https://github.com/malbouis/Python_intro/blob/master/aulas_2019/pics/ref_list.png?raw=1)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64zoncRQlfAi"
      },
      "source": [
        "a=[1,2,3]\n",
        "b=[1,2,3]\n",
        "a is b"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2HNrtZH7ok_Q"
      },
      "source": [
        "## Alias\n",
        "\n",
        "Podemos forçar ***a*** e ***b*** a se referirem ao mesmo objeto. Com isso, criamos um **alias** (apelido).\n",
        "\n",
        "No exemplo acima, **a** e **b** são objetos distintos."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KYAxoh4kiusN"
      },
      "source": [
        "a is b"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mcE0QCU6oz3s"
      },
      "source": [
        "\n",
        "Entretanto, podemos forçar **a** e **b** se referirem ao mesmo objeto, criando um **alias**, da seguinte maneira:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O8nefcShiusP"
      },
      "source": [
        "a = b\n",
        "a is b"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tZr-gar2iusR"
      },
      "source": [
        "a.append(88)\n",
        "print('a = ', a, ', b = ', b)\n",
        "# b = ?\n",
        "# a = ?"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "30ESoTdYiusT"
      },
      "source": [
        "### O que aconteceu???\n",
        "\n",
        "![difflists](https://github.com/malbouis/Python_intro/blob/master/aulas_2019/pics/diff_lists.png?raw=1)\n",
        "![samelists](https://github.com/malbouis/Python_intro/blob/master/aulas_2019/pics/same_lists.png?raw=1)\n",
        "\n",
        "\n",
        "### Recomendações em relação a **aliases**:\n",
        "Em geral, é mais seguro evitar fazer um **alias** de objetos **mutáveis**, como listas. Para objetos **imutáveis**, como strings, não há restrições de se fazer **aliases**.\n",
        "\n",
        "## Clonagem de listas\n",
        "\n",
        "Se quisermos modificar uma lista e ao mesmo tempo mantermos uma cópia inalterada dessa lista, devemos **cloná-la**:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FQP0rFatiusU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "29b991bf-28ac-4815-9e0e-24f3e80b7455"
      },
      "source": [
        "lista = [1, 2, 3]\n",
        "clone_lista = lista[:] # clone feito ao tomar uma fatia de toda a lista.\n",
        "clone_lista"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1, 2, 3]"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eJWj0pKliusV"
      },
      "source": [
        "Podemos então criar uma nova lista ao tomar uma **fatia** de **toda a lista**.\n",
        "\n",
        "![reflist](https://github.com/malbouis/Python_intro/blob/master/aulas_2019/pics/ref_list.png?raw=1)\n",
        "\n",
        "\n",
        "E agora, como cada uma das listas é uma referência a um objeto diferente, podemos modificar uma variável, sem modificar a outra:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5pinnZzAiusW",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7353741d-ad0d-49c5-98cf-9c361a2ed764"
      },
      "source": [
        "a = [1, 2, 3]\n",
        "b = a[:]\n",
        "a.append(88)\n",
        "print ('a: ', a, ', b: ', b)\n",
        "\n",
        "b[2] = 199\n",
        "print ('a: ', a, ', b: ', b)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "a:  [1, 2, 3, 88] , b:  [1, 2, 3]\n",
            "a:  [1, 2, 3, 88] , b:  [1, 2, 199]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Mj-Yyw9KiusY"
      },
      "source": [
        "## Listas e loops ***for***\n",
        "\n",
        "Suponha que temos duas listas. Se elas forem de tamanhos iguais, queremos imprimir seus valores em um único loop."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C3UMA_oZiusY"
      },
      "source": [
        "lista1 = [1,3,5,7,9,11,13,15]\n",
        "lista2 = [2,4,6,8,10,12,14,16]\n",
        "\n",
        "# as listas têm o mesmo tamanho?\n",
        "len(lista1) == len(lista2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "98grbrQyiusb"
      },
      "source": [
        "# se as listas têm o mesmo número de elementos, imprima todos os elementos de cada uma das listas\n",
        "if (len(lista1) == len(lista2)):\n",
        "    print(range(len(lista1)))\n",
        "    for i in range(len(lista1)):\n",
        "        print(lista1[i])\n",
        "        print(lista2[i])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ea_39iPgmZMA"
      },
      "source": [
        "Podemos usar a função ``enumerate`` para percorrer uma lista e também obter os índices:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mo9rE7iAiuse"
      },
      "source": [
        "alunos = ['Ana', 'Luiz', 'Maria', 'Joana']\n",
        "idades = [16, 18, 17, 19]\n",
        "\n",
        "if (len(alunos)==len(idades)):\n",
        "    for (i, valor) in enumerate(alunos):\n",
        "        print(valor, ' tem ', idades[i], ' anos')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iPl2pO-Jiush"
      },
      "source": [
        "## Métodos de listas\n",
        "\n",
        "As listas são um tipo de dados de python. Sendo assim têm métodos (funções) associadas a elas, que são acessadas através do ponto:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eibBqsJ2iusj"
      },
      "source": [
        "mylist = []\n",
        "mylist.append(5)\n",
        "mylist.append(27)\n",
        "mylist.append(3)\n",
        "mylist.append(12)\n",
        "mylist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R3BI2G9Ziusl"
      },
      "source": [
        "mylist.insert(1, 12) # insere o item 12 na posição 1. Todos os outros itens da lista são desolcados para cima.\n",
        "mylist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2-8y2Y7ciusr"
      },
      "source": [
        "mylist.count(12) # conta quantas vezes o valor 12 aparece na lista"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E2Yfop5Riust"
      },
      "source": [
        "mylist.extend([5, 9, 5, 11])  # coloca toda a lista do argumento no final da lista mylist\n",
        "mylist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U2atvTmaiusu"
      },
      "source": [
        "mylist.index(9) # encontra o índice do primeiro valor 9 de mylist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S7ZmWv2hiusx"
      },
      "source": [
        "mylist.reverse() # reverte todos os objetos de mylist\n",
        "mylist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VDAoU1A8ius0"
      },
      "source": [
        "mylist.sort()  # ordena os valores de uma lista\n",
        "mylist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eeV5Dvyjius4"
      },
      "source": [
        "mylist.remove(12)  # remove o primeiro elemento de valor 12 de mylist\n",
        "mylist"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "miEvQWoiius6"
      },
      "source": [
        "## Strings e listas\n",
        "\n",
        "Dois dos métodos mais úteis em ***strings*** envolvem conversão de ***strings*** para ***listas*** de substrings e vice-versa. \n",
        "\n",
        "\n",
        "O método ***split*** divide uma ***string*** em uma ***lista*** de palavras. O padrão é considerar caracteres de espaço em branco como **delimitador** de palavras:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7V0AYKrzius6"
      },
      "source": [
        "musica = \"O sol há de brilhar mais uma vez\"\n",
        "palavras = musica.split()\n",
        "palavras"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wyl93fYwius9"
      },
      "source": [
        "Podemos especificar um **delimitador**. Por exemplo, podemos escolher a letra 'a' como **delimitador**, no lugar do padrão."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OI3L5jBcius-"
      },
      "source": [
        "palavras = musica.split('a')\n",
        "palavras"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fqQwoO6GiutD"
      },
      "source": [
        "O inverso do método ***split*** é o ***join***. Para juntar as strings, escolhemos um **separador** como parâmetro."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4jvoAPmZiutG"
      },
      "source": [
        "cola = 'a'\n",
        "palavras_coladas = cola.join(palavras)\n",
        "palavras_coladas"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hwq5E61_iutI"
      },
      "source": [
        "## List Comprehension\n",
        "\n",
        "A compreensão de listas fornece uma maneira concisa de criar listas. \n",
        "Aplicações comuns são fazer novas listas onde cada elemento é o resultado de algumas operações aplicadas a cada membro de outra seqüência ou iterável, ou criar uma subsequência daqueles elementos que satisfazem uma determinada condição."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZsOWODOoiutJ"
      },
      "source": [
        "squares = []\n",
        "for x in range(10):\n",
        "    squares.append(x**2)\n",
        "\n",
        "print(squares)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "blPuBvJxiutL"
      },
      "source": [
        "squares2 = [x**2 for x in range(10)]\n",
        "print(squares2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z8OL13GviutP"
      },
      "source": [
        "[(x,y) for x in [1,2,3] for y in [3,1,4] if x != y]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e2wFrReeiutU"
      },
      "source": [
        "from math import pi\n",
        "[str(round(pi, i)) for i in range(1, 6)]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E9ITlrNBiutZ"
      },
      "source": [
        "## Exercícios\n",
        "\n",
        "1) Considere a função abaixo. Esse tipo de função é considerada uma **função modificadora** pois modifica o objeto que é passado como argumento. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iKwmsYgeiutZ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "91e4be0a-168e-4be5-b01e-1244f42d1de8"
      },
      "source": [
        "def dobrar_elementos(uma_lista):\n",
        "    \"\"\" Retorma uma 'nova_lista' que reescreve os elementos de 'uma_lista' com o dobro de seus valores originais, sem modificar os valores de 'uma_lista'.\n",
        "    \"\"\"\n",
        "    nova_lista = uma_lista[:]\n",
        "    for (i, valor) in enumerate(uma_lista):\n",
        "        novo_elem = 2 * valor\n",
        "        nova_lista[i] = novo_elem\n",
        "\n",
        "    return nova_lista\n",
        "\n",
        "minha_lista = [2, 4, 6]\n",
        "print(minha_lista)\n",
        "lista_dobrada = dobrar_elementos(minha_lista)\n",
        "print(lista_dobrada)\n",
        "print(minha_lista)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[2, 4, 6]\n",
            "[4, 8, 12]\n",
            "[2, 4, 6]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i23JYSmRiutb"
      },
      "source": [
        "a) Modifique a função para retornar uma **nova lista**, sem modificar a lista usada como parâmetro. Esse tipo de função é chamado de **função pura**.\n",
        "\n",
        "\n",
        "b) Modifique a documentação de ajuda da nova função, de tal forma que quando se chame a função ***help*** da nova função, se obtenha a descrição adequada."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RgYPzOGCiutc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "874e74d0-8fff-4cc7-c3b7-87f435687b07"
      },
      "source": [
        "help(dobrar_elementos)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Help on function dobrar_elementos in module __main__:\n",
            "\n",
            "dobrar_elementos(uma_lista)\n",
            "    Retorma uma 'nova_lista' que reescreve os elementos de 'uma_lista' com o dobro de seus valores originais, sem modificar os valores de 'uma_lista'.\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tmKDZTws-PuD"
      },
      "source": [
        "# sorteando elementos de uma lista com o módulo random:\n",
        "import random\n",
        "\n",
        "nomes=['Helena', 'Dilson', 'Janaina', 'Isabela']\n",
        "sorteado = random.choice(nomes)\n",
        "print(sorteado)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GYSWbfQ33TrD"
      },
      "source": [
        "# Módulos e Arquivos\n",
        "\n",
        "\n",
        "## Módulos\n",
        "\n",
        "Um módulo é um arquivo contendo definições e instruções do Python destinadas a uso em ***outros programas*** do Python. Existem muitos módulos de Python que vêm como parte da biblioteca padrão. Já vimos (pelo menos) dois deles, o módulo ```turtle``` e o módulo ```string```. Documentação sobre esses e outros módulos pode ser encontrada no site https://docs.python.org/3.5/library/index.html.\n",
        "\n",
        "\n",
        "**Nessa aula mostraremos como qualquer arquivo que contenha código Python pode ser importado como um módulo.**\n",
        "\n",
        "### Criando módulos \n",
        "Tudo o que precisamos para criar um módulo é um arquivo de texto com uma extensão .py no nome do arquivo:\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tga6ov2i3TrE"
      },
      "source": [
        "# esse comando logo abaixo é do jupyter notebook para escrever um arquivo .py\n",
        "%%writefile Seqtools.py \n",
        "def  remover_em ( pos ,  seq ): \n",
        "    return  seq [: pos ]  +  seq [ pos + 1 :]\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3DxthzW2H1aC"
      },
      "source": [
        "# ler o aqruivo que acabamos de criar:\n",
        "%cat Seqtools.py"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BNuaN-QX3TrI"
      },
      "source": [
        "Agora podemos usar nosso módulo nos scripts e no shell do Python. Para fazer isso, devemos primeiro importar o módulo. Existem duas maneiras de fazer isso:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GaFnnDWI3TrJ"
      },
      "source": [
        "import  Seqtools \n",
        "s  =  \"O string!\" \n",
        "Seqtools.remover_em( 4 ,  s ) "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sP10o5r93TrO"
      },
      "source": [
        "ou também"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LjGUT1nr3TrP"
      },
      "source": [
        "from Seqtools import remover_em\n",
        "s = \"o string!\"\n",
        "remover_em(4,s)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V9_jXhY53TrS"
      },
      "source": [
        "No primeiro exemplo o nome do módulo e um ponto é escrito antes do nome da função, separados de um ponto(```.```). No segundo exemplo só a função ```remover_em``` é importada, e a  chamada é feita exatamente como as funções que vimos anteriormente, sem necesidade de incluir o nome do módulo.\n",
        "\n",
        "Observe que, em ambos os casos, não incluímos a extensão do arquivo ```.py``` ao importar. O Python espera que os nomes de arquivos dos módulos do Python sejam finalizados em ```.py``` , portanto, a extensão do arquivo não é incluída na instrução de importação.\n",
        "\n",
        "O uso de módulos permite **dividir programas muito grandes** em partes de **tamanho gerenciável** e manter as partes relacionadas juntas."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gCVmHeel3TrT"
      },
      "source": [
        "Outro exemplo é apresentado com  o arquivo ```lcount.py```  definido assim:\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sn2hs_YeJj8f"
      },
      "source": [
        "%%writefile lcount.py\n",
        "\n",
        "def linecount(filename):\n",
        "    count = 0\n",
        "    for line in open(filename):\n",
        "        count += 1\n",
        "    return count\n",
        "\n",
        "linecount('lcount.py')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MCXPlsE7JxRR"
      },
      "source": [
        "%cat lcount.py"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jcM3SMT0KVYi"
      },
      "source": [
        "linecount('lcount.py')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "d65YS7gE3TrY"
      },
      "source": [
        "Se você executar este programa, ele lê a si mesmo e imprime o número de linhas no arquivo, que é 7. \n",
        "\n",
        "Você também pode importá-lo:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H0TOZOzF3TrZ"
      },
      "source": [
        "import lcount"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kmc4yJLc3Trc"
      },
      "source": [
        "Agora podemos usar a função do módulo em outros programas, pois temos o **objeto** `lcount` com sua função associada."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1VhucDFu3Trd"
      },
      "source": [
        "print(lcount.linecount(\"Seqtools.py\"))\n",
        "#print(lcount.linecount(\"dados_alunos.txt\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oEIlmEg13Trg"
      },
      "source": [
        "O único problema com este exemplo é que quando você importa o módulo, ele executa o código de teste na parte inferior. Normalmente, quando você importa um módulo, ele define novas funções, mas não as executa.\n",
        "\n",
        "Programas que serão importados como módulos geralmente usam o seguinte idioma:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hrW5yWZe3Trh"
      },
      "source": [
        "if __name__ == '__main__':\n",
        "    print(linecount('lcount.py'))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pWryVMSB3Trk"
      },
      "source": [
        "Pontos importantes:\n",
        "\n",
        "* O identificador ```__name__``` é uma variável interna que é configurada quando o programa é iniciado; \n",
        "* Se o programa estiver sendo executado como um script, ```__name__``` tem o valor ```'__main__'```; \n",
        "   * Nesse caso, o código de teste é executado. \n",
        "\n",
        "* Caso contrário, se o módulo estiver sendo ***importado***, o código de teste será ignorado."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JrWQHOZA3Trk"
      },
      "source": [
        "## Arquivos\n",
        "\n",
        "Este capítulo introduz a ideia de programas **“persistentes”** que mantêm os dados em armazenamento permanente e mostra como usar diferentes tipos de armazenamento permanente, como arquivos.\n",
        "\n",
        "\n",
        "### Persistência\n",
        "\n",
        "* A maioria dos programas que vimos até agora é **transitória**:\n",
        "   * são executados por um curto período de tempo e produzem alguma saída, mas quando terminam, seus dados desaparecem. \n",
        "   * Se você executar o programa novamente, ele começa do zero.\n",
        "\n",
        "* Outros programas são **persistentes**: \n",
        "   * são executados por um longo período (ou o tempo todo); \n",
        "   * mantêm pelo menos alguns dos seus dados em armazenamento permanente (um disco rígido, por exemplo); \n",
        "   * se eles desligarem e reiniciarem, eles continuam de onde pararam.\n",
        "   \n",
        "\n",
        "* Uma das formas mais simples de programas manterem seus dados é lendo e gravando arquivos de texto.\n",
        "* Outra alternativa é o uso de banco de dados, porém não abordaremos esse assunto aqui.\n",
        "\n",
        "\n",
        "## Leitura e escrita de arquivos\n",
        "\n",
        "Um arquivo de texto é uma sequência de caracteres armazenados em um meio permanente, como um disco rígido, uma memória flash ou um CD-ROM. \n",
        "\n",
        "### Leitura de um arquivo de texto\n",
        "Vamos criar um arquivo, 'dados_alunos.txt'.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kpb-ypaThrJV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ca0c5f20-edc1-4e08-f086-c35c5c79a82d"
      },
      "source": [
        "%%writefile dados_alunos.txt\n",
        "18\t1.68\t80\n",
        "18\t1.94\t60\n",
        "18\t1.7\t80\n",
        "18\t1.76\t66\n",
        "19\t1.73\t87.5\n",
        "18\t1.66\t58\n",
        "21\t1.8\t92\n",
        "18\t1.6\t57\n",
        "18\t1.67\t64\n",
        "18\t1.73\t57\n",
        "17\t1.73\t75\n",
        "18\t1.61\t59\n",
        "18\t1.69\t90\n",
        "17\t1.71\t67\n",
        "19\t1.78\t60\n",
        "22\t1.68\t72\n",
        "18\t1.7\t73\n",
        "19\t1.64\t86\n",
        "18\t1.64\t75\n",
        "20\t1.8\t95\n",
        "17\t1.75\t60\n",
        "18\t1.78\t75\n",
        "18\t1.75\t65\n",
        "17\t1.69\t60\n",
        "19\t1.78\t73\n",
        "18\t1.7\t63\n",
        "34\t1.75\t78\n",
        "18\t1.64\t64\n",
        "19\t1.75\t50\n",
        "18\t1.67\t61\n",
        "18\t1.7\t70\n",
        "20\t1.8\t60\n",
        "18\t1.63\t57\n",
        "23\t1.89\t110\n",
        "18\t1.71\t71\n",
        "18\t1.65\t65\n",
        "17\t1.72\t67\n",
        "19\t1.65\t58\n",
        "18\t1.75\t90\n",
        "18\t1.7\t64\n",
        "19\t1.81\t70\n",
        "19\t1.65\t43\n",
        "28\t1.52\t50\n",
        "19\t1.79\t78\n",
        "26\t1.79\t82\n",
        "19\t1.75\t61\n",
        "19\t1.8\t70\n",
        "20\t1.75\t70\n",
        "20\t1.73\t70\n",
        "19\t1.7\t50\n",
        "22\t1.78\t72\n",
        "19\t1.77\t55\n",
        "18\t1.53\t58\n",
        "28\t1.54\t50\n",
        "20\t1.83\t70\n",
        "44\t1.85\t90\n",
        "18\t1.6\t51"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Writing dados_alunos.txt\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HODFJmTgh4ec",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2406b312-e8d0-4e0a-ef5b-6438f5ab537b"
      },
      "source": [
        "%cat dados_alunos.txt"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "18\t1.68\t80\n",
            "18\t1.94\t60\n",
            "18\t1.7\t80\n",
            "18\t1.76\t66\n",
            "19\t1.73\t87.5\n",
            "18\t1.66\t58\n",
            "21\t1.8\t92\n",
            "18\t1.6\t57\n",
            "18\t1.67\t64\n",
            "18\t1.73\t57\n",
            "17\t1.73\t75\n",
            "18\t1.61\t59\n",
            "18\t1.69\t90\n",
            "17\t1.71\t67\n",
            "19\t1.78\t60\n",
            "22\t1.68\t72\n",
            "18\t1.7\t73\n",
            "19\t1.64\t86\n",
            "18\t1.64\t75\n",
            "20\t1.8\t95\n",
            "17\t1.75\t60\n",
            "18\t1.78\t75\n",
            "18\t1.75\t65\n",
            "17\t1.69\t60\n",
            "19\t1.78\t73\n",
            "18\t1.7\t63\n",
            "34\t1.75\t78\n",
            "18\t1.64\t64\n",
            "19\t1.75\t50\n",
            "18\t1.67\t61\n",
            "18\t1.7\t70\n",
            "20\t1.8\t60\n",
            "18\t1.63\t57\n",
            "23\t1.89\t110\n",
            "18\t1.71\t71\n",
            "18\t1.65\t65\n",
            "17\t1.72\t67\n",
            "19\t1.65\t58\n",
            "18\t1.75\t90\n",
            "18\t1.7\t64\n",
            "19\t1.81\t70\n",
            "19\t1.65\t43\n",
            "28\t1.52\t50\n",
            "19\t1.79\t78\n",
            "26\t1.79\t82\n",
            "19\t1.75\t61\n",
            "19\t1.8\t70\n",
            "20\t1.75\t70\n",
            "20\t1.73\t70\n",
            "19\t1.7\t50\n",
            "22\t1.78\t72\n",
            "19\t1.77\t55\n",
            "18\t1.53\t58\n",
            "28\t1.54\t50\n",
            "20\t1.83\t70\n",
            "44\t1.85\t90\n",
            "18\t1.6\t51"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rc6SplvJhpGL"
      },
      "source": [
        "Para ler o arquivo, usamos a seguinte sintaxe:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r3uLi_1D3Trl",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "3e757a00-3b22-47b5-c674-3ee622880554"
      },
      "source": [
        "fin = open('dados_alunos.txt')\n",
        "fin.readline()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'18\\t1.68\\t80\\n'"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t63lEwtZ3Tro"
      },
      "source": [
        "O método readline() se atualiza para cada linha lida:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cWDLZa9-3Tro",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "0bc10a60-9009-4c07-fbc3-39158ba805fd"
      },
      "source": [
        "fin.readline()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'18\\t1.94\\t60\\n'"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9Q7jWtEw3Trr",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "53089ee2-6327-4304-895a-551dab01eb04"
      },
      "source": [
        "linha = fin.readline()\n",
        "linha.strip()"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'18\\t1.7\\t80'"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UzjEpw9E3Tru"
      },
      "source": [
        "##### Lendo um arquivo com várias colunas:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "us6ggEPu3Tru",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "380fc934-d811-4000-ef3c-ee1d3b6f471b"
      },
      "source": [
        "fin = open('dados_alunos.txt')\n",
        "linhas = fin.readlines()\n",
        "print (linhas)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['18\\t1.68\\t80\\n', '18\\t1.94\\t60\\n', '18\\t1.7\\t80\\n', '18\\t1.76\\t66\\n', '19\\t1.73\\t87.5\\n', '18\\t1.66\\t58\\n', '21\\t1.8\\t92\\n', '18\\t1.6\\t57\\n', '18\\t1.67\\t64\\n', '18\\t1.73\\t57\\n', '17\\t1.73\\t75\\n', '18\\t1.61\\t59\\n', '18\\t1.69\\t90\\n', '17\\t1.71\\t67\\n', '19\\t1.78\\t60\\n', '22\\t1.68\\t72\\n', '18\\t1.7\\t73\\n', '19\\t1.64\\t86\\n', '18\\t1.64\\t75\\n', '20\\t1.8\\t95\\n', '17\\t1.75\\t60\\n', '18\\t1.78\\t75\\n', '18\\t1.75\\t65\\n', '17\\t1.69\\t60\\n', '19\\t1.78\\t73\\n', '18\\t1.7\\t63\\n', '34\\t1.75\\t78\\n', '18\\t1.64\\t64\\n', '19\\t1.75\\t50\\n', '18\\t1.67\\t61\\n', '18\\t1.7\\t70\\n', '20\\t1.8\\t60\\n', '18\\t1.63\\t57\\n', '23\\t1.89\\t110\\n', '18\\t1.71\\t71\\n', '18\\t1.65\\t65\\n', '17\\t1.72\\t67\\n', '19\\t1.65\\t58\\n', '18\\t1.75\\t90\\n', '18\\t1.7\\t64\\n', '19\\t1.81\\t70\\n', '19\\t1.65\\t43\\n', '28\\t1.52\\t50\\n', '19\\t1.79\\t78\\n', '26\\t1.79\\t82\\n', '19\\t1.75\\t61\\n', '19\\t1.8\\t70\\n', '20\\t1.75\\t70\\n', '20\\t1.73\\t70\\n', '19\\t1.7\\t50\\n', '22\\t1.78\\t72\\n', '19\\t1.77\\t55\\n', '18\\t1.53\\t58\\n', '28\\t1.54\\t50\\n', '20\\t1.83\\t70\\n', '44\\t1.85\\t90\\n', '18\\t1.6\\t51']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4FbvpMGX3Tr0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cb454486-ae0c-4bbd-efbc-d7d4b77b097e"
      },
      "source": [
        "for line in linhas:\n",
        "    #print (line)\n",
        "    column = line.strip().split('\\t')\n",
        "    print (column)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['18', '1.68', '80']\n",
            "['18', '1.94', '60']\n",
            "['18', '1.7', '80']\n",
            "['18', '1.76', '66']\n",
            "['19', '1.73', '87.5']\n",
            "['18', '1.66', '58']\n",
            "['21', '1.8', '92']\n",
            "['18', '1.6', '57']\n",
            "['18', '1.67', '64']\n",
            "['18', '1.73', '57']\n",
            "['17', '1.73', '75']\n",
            "['18', '1.61', '59']\n",
            "['18', '1.69', '90']\n",
            "['17', '1.71', '67']\n",
            "['19', '1.78', '60']\n",
            "['22', '1.68', '72']\n",
            "['18', '1.7', '73']\n",
            "['19', '1.64', '86']\n",
            "['18', '1.64', '75']\n",
            "['20', '1.8', '95']\n",
            "['17', '1.75', '60']\n",
            "['18', '1.78', '75']\n",
            "['18', '1.75', '65']\n",
            "['17', '1.69', '60']\n",
            "['19', '1.78', '73']\n",
            "['18', '1.7', '63']\n",
            "['34', '1.75', '78']\n",
            "['18', '1.64', '64']\n",
            "['19', '1.75', '50']\n",
            "['18', '1.67', '61']\n",
            "['18', '1.7', '70']\n",
            "['20', '1.8', '60']\n",
            "['18', '1.63', '57']\n",
            "['23', '1.89', '110']\n",
            "['18', '1.71', '71']\n",
            "['18', '1.65', '65']\n",
            "['17', '1.72', '67']\n",
            "['19', '1.65', '58']\n",
            "['18', '1.75', '90']\n",
            "['18', '1.7', '64']\n",
            "['19', '1.81', '70']\n",
            "['19', '1.65', '43']\n",
            "['28', '1.52', '50']\n",
            "['19', '1.79', '78']\n",
            "['26', '1.79', '82']\n",
            "['19', '1.75', '61']\n",
            "['19', '1.8', '70']\n",
            "['20', '1.75', '70']\n",
            "['20', '1.73', '70']\n",
            "['19', '1.7', '50']\n",
            "['22', '1.78', '72']\n",
            "['19', '1.77', '55']\n",
            "['18', '1.53', '58']\n",
            "['28', '1.54', '50']\n",
            "['20', '1.83', '70']\n",
            "['44', '1.85', '90']\n",
            "['18', '1.6', '51']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kcWZv6QTjzww"
      },
      "source": [
        "### O módulo Pandas\n",
        "\n",
        "Também é possível ler e escrever arquivos através do módulo Pandas.\n",
        "\n",
        "Vamos criar um arquivo com cabeçalho, baseado no arquivo `dados_alunos.txt` e lê-lo com o `pandas`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XA78uB3-kQqb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c2aee010-55e6-405a-c529-c95307fc5954"
      },
      "source": [
        "%%writefile dados_alunos_cabecalho.csv\n",
        "I(a),A(m),M(Kg)\n",
        "18,1.68,80\n",
        "18,1.94,60\n",
        "18,1.7,80\n",
        "18,1.76,66\n",
        "19,1.73,87.5\n",
        "18,1.66,58\n",
        "21,1.8,92\n",
        "18,1.6,57\n",
        "18,1.67,64\n",
        "18,1.73,57"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Writing dados_alunos_cabecalho.csv\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7-15kiPYiTlJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "e574ae2c-b259-4fbb-f7bd-d7009ce9d605"
      },
      "source": [
        "import pandas as pd\n",
        "\n",
        "df = pd.read_csv('dados_alunos_cabecalho.csv')\n",
        "\n",
        "df.head()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>I(a)</th>\n",
              "      <th>A(m)</th>\n",
              "      <th>M(Kg)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>18</td>\n",
              "      <td>1.68</td>\n",
              "      <td>80.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>1.94</td>\n",
              "      <td>60.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>18</td>\n",
              "      <td>1.70</td>\n",
              "      <td>80.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>18</td>\n",
              "      <td>1.76</td>\n",
              "      <td>66.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>19</td>\n",
              "      <td>1.73</td>\n",
              "      <td>87.5</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   I(a)  A(m)  M(Kg)\n",
              "0    18  1.68   80.0\n",
              "1    18  1.94   60.0\n",
              "2    18  1.70   80.0\n",
              "3    18  1.76   66.0\n",
              "4    19  1.73   87.5"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DrRedbeEnW_J",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "35c80c23-41c3-4830-c038-9ad5e4f3e502"
      },
      "source": [
        "df.tail()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>I(a)</th>\n",
              "      <th>A(m)</th>\n",
              "      <th>M(Kg)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>18</td>\n",
              "      <td>1.66</td>\n",
              "      <td>58.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>21</td>\n",
              "      <td>1.80</td>\n",
              "      <td>92.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>18</td>\n",
              "      <td>1.60</td>\n",
              "      <td>57.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>18</td>\n",
              "      <td>1.67</td>\n",
              "      <td>64.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>18</td>\n",
              "      <td>1.73</td>\n",
              "      <td>57.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   I(a)  A(m)  M(Kg)\n",
              "5    18  1.66   58.0\n",
              "6    21  1.80   92.0\n",
              "7    18  1.60   57.0\n",
              "8    18  1.67   64.0\n",
              "9    18  1.73   57.0"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kpHFSiQlnb2i",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "89844274-181b-4727-bf6b-e6b10c367657"
      },
      "source": [
        "df['I(a)']"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    18\n",
              "1    18\n",
              "2    18\n",
              "3    18\n",
              "4    19\n",
              "5    18\n",
              "6    21\n",
              "7    18\n",
              "8    18\n",
              "9    18\n",
              "Name: I(a), dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jiJ-X6uX3Tr3"
      },
      "source": [
        "## Escrita de um arquivo de texto\n",
        "\n",
        "Para escrever um arquivo, você precisa abri-lo com o modo 'w' como segundo parâmetro:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SqMmQJbP3Tr4"
      },
      "source": [
        "fw = open('output.txt', 'w')"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nIZ8h20y3Tr6"
      },
      "source": [
        "* Se o arquivo já existir e abrí-lo no modo de escrita, os dados antigos são apagados e a escrita começa de novo. \n",
        "* Se o arquivo não existir, um novo será criado.\n",
        "\n",
        "\n",
        "* ***open*** retorna um objeto de arquivo que fornece métodos para trabalhar com o arquivo. \n",
        "* O método ***write*** coloca dados no arquivo."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aGHtiQrz3Tr7"
      },
      "source": [
        "linha1 = fw.write('primeira linha do arquivo. \\n')\n",
        "linha2 = fw.write('segunda linha do arquivo. \\n')\n",
        "print(linha1,linha2) # linha1 e linha2 salvam o numero de carateres escritos\n",
        "fw.close()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cbm3F-2hTdzJ"
      },
      "source": [
        "fw = open('output.txt', 'r')\n",
        "fw.readlines()\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0vauasHg3Tr9"
      },
      "source": [
        "## Operador de formatação\n",
        "\n",
        "O argumento do método ***write*** deve ser uma string, então, para escrever valores em um arquivo, temos que **convertê-los em strings**. A maneira mais fácil de fazer isso é com ***str***."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A8sU3jgr3Tr9"
      },
      "source": [
        "fout = open('ex_aula13.txt', 'w')\n",
        "x = 52\n",
        "fout.write(str(x))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NSee7vAh3Tr_"
      },
      "source": [
        "Uma alternativa é usar o operador de formatação,%. \n",
        "   * Quando aplicado a números inteiros,% é o operador de módulo. \n",
        "   * Quando o primeiro operando é uma string,% é o operador de formatação.\n",
        "\n",
        "\n",
        "Por exemplo, a seqüência de formato **'%d'** significa que o segundo operando deve ser formatado como um **inteiro decimal**:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5-hD2I6v3TsA"
      },
      "source": [
        "camelos = 42\n",
        "'%d' % camelos\n",
        "#'42'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BvAmaEzx3TsC"
      },
      "source": [
        "Uma seqüência de formato pode aparecer em qualquer lugar da string, então você pode inserir um valor em uma frase:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hmvLFoRe3TsC"
      },
      "source": [
        "'Temos ainda pelo menos %i projetos antes do final de %4.1f' % (2, 2020.1) "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Cu2EPkEk3TsE"
      },
      "source": [
        "#### Outra forma de formatação é usando o método de string 'format':"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KAxXBWgi3TsF"
      },
      "source": [
        "'Temos ainda pelo menos {0} projetos antes do final de {1:4.1f}'.format(2, 2020.1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OXcJfsnY3TsH"
      },
      "source": [
        "#### Novo estilo de formatação : f-strings ou formatted string litterals\n",
        "https://docs.python.org/3/reference/lexical_analysis.html#f-strings\n",
        "\n",
        "Uma forma mais \"moderna\" de formatar é utilizando as chamadas f-strings, que foi incluído a partir da versão do Python 3.6.\n",
        "\n",
        "* f-strings são prefixados com 'f' e são semelhantes às strings de formatação aceitas por `str.format()`;\n",
        "* Elas contêm campos de substituição entre chaves;\n",
        "   * Os campos de substituição são expressões, que são avaliadas em tempo de execução e, em seguida, formatadas usando o protocolo `format()`.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ic1QzCTB3TsH"
      },
      "source": [
        "f\"Tenho {2} projetos antes do final de {2020.2:4.1f} \""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PapfzolC3TsK"
      },
      "source": [
        "f\"Os camelos sabem o sentido da vida {camelos}\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OlZuCWtA3TsM"
      },
      "source": [
        "## Nomes de arquivo e *paths*\n",
        "\n",
        "Através do módulo ***os*** podemos ter acesso a informação sobre diretórios e arquivos: "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "maPVOMiD3TsN"
      },
      "source": [
        "import os\n",
        "cwd = os.getcwd()    #current working directory (diretório de trabalho)\n",
        "cwd"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4frjEChI3TsQ"
      },
      "source": [
        "O módulo ***os*** fornece vários métodos úteis de verificação de ***path*** e listagem de arquivos em um dado diretório:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B8sOuli63TsQ"
      },
      "source": [
        "os.path.abspath('ex_aula13.txt')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hDF7bbHt3TsT"
      },
      "source": [
        "os.path.exists('ex_aula13.txt')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yNdrg5E43TsX"
      },
      "source": [
        "os.path.isdir('ex_aula13.txt')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kRVlBE_k3TsZ"
      },
      "source": [
        "os.path.isdir('/Users/mariaclemenciamoraherrera/cernbox/PCUERJ/PythonUERJ/Python_intro_github/aulas/')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PZXJJQNT3Tsc"
      },
      "source": [
        "os.path.isfile('ex_aula13.txt')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z6QIYA5f3Tse"
      },
      "source": [
        "os.listdir(cwd)   # lista os arquivos e subdiretórios do diretório atual"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NV54e51P3Tsg"
      },
      "source": [
        "## Capturando exceções\n",
        "\n",
        "Há formas de prever e lidar com erros e exceções. \n",
        "\n",
        "Alguns exemplos de exceções:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hkyfu9j_3Tsh"
      },
      "source": [
        "fin = open('bad_file')    # quando o arquivo não existe"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RdGkSUXH3Tsj"
      },
      "source": [
        "fout = open('/etc/passwd', 'w')  # quando não temos permissão de escrever no diretório"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WeK2vjNF3Tsl"
      },
      "source": [
        "fin = open('/home')  # tentativa de abrir um diretório para leitura"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KCfEAFHN3Tsn"
      },
      "source": [
        "**Para evitar esses erros, você poderia usar funções como *os.path.exists* e *os.path.isfile*, mas levaria muito tempo e código para verificar todas as possibilidades (se “Errno 21” for qualquer indicação, há pelo menos 21 coisas que podem dar errado).**\n",
        "\n",
        "É melhor fazer uma tentativa e lidar com possíveis problemas. A declaração **try** faz isso!!!!"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Smph0R-93Tsn"
      },
      "source": [
        "try:    \n",
        "    fin = open('bad_file')\n",
        "except:\n",
        "    print('Algo de errado aconteceu.')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0GgAFj8C3Tsp"
      },
      "source": [
        "* O Python começa executando a cláusula ***try***. \n",
        "   * Se tudo correr bem:\n",
        "      * ele pula a cláusula ***except*** e continua. \n",
        "   * Se ocorrer uma exceção: \n",
        "      * ele sai da cláusula ***try*** e executa a cláusula ***except***.\n",
        "      \n",
        "\n",
        "* Manipular uma exceção com uma instrução ***try*** é chamado de **captura de uma exceção**. \n",
        "\n",
        "Para leitura sobre tipos de erro e a sua manipulação, sugerimos o tutorial https://docs.python.org/3/tutorial/errors.html\n",
        "\n",
        "\n",
        "Uma forma de obter mais informação sobre o erro é chamar a excepção  com um identificador (nesse caso ```err```) e imprimi-lo assim como seu tipo."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "miZv8Z3C3Tsr"
      },
      "source": [
        "try:\n",
        "    fin = open(\"another_bad_file\")\n",
        "except Exception as err:\n",
        "    print(type(err))\n",
        "    print(err)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nkLDFzFC3Tst"
      },
      "source": [
        "## Tubos (pipes)\n",
        "\n",
        "A maioria dos sistemas operacionais fornece uma **interface de linha de comando**, também conhecida como **shell**. As shells geralmente fornecem comandos para navegar no sistema de arquivos e iniciar aplicativos.\n",
        "\n",
        "\n",
        "Qualquer programa que você pode iniciar a partir da shell também pode ser iniciado a partir do Python usando um objeto **pipe**, que representa um programa em execução.\n",
        "\n",
        "Por exemplo, o comando Unix *ls -l* exibe o conteúdo do diretório atual em formato longo. Você pode iniciar o *ls* com ***os.popen***:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3rGU836h3Tst"
      },
      "source": [
        "cmd = 'ls -lh'\n",
        "fp = os.popen(cmd)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V1Uhj1sj3Tsv"
      },
      "source": [
        "* O argumento é uma string que contém um comando shell. \n",
        "* O valor de retorno é um objeto que se comporta como um **arquivo aberto**. \n",
        "* Você pode ler a saída do processo *ls* uma linha de cada vez com ***readline*** ou obter tudo de uma vez com ***read***:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dPhwoS1G3Tsw"
      },
      "source": [
        "print(fp.read())"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DsbuFVEG3Tsy"
      },
      "source": [
        "stat = fp.close()\n",
        "print(stat)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oJpfbRJQ3Ts0"
      },
      "source": [
        "O valor de retorno é o status final do processo ```ls -lh```; ***None*** significa que terminou normalmente (sem erros)."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LkKV2_H23Ts1"
      },
      "source": [
        "## Exercícios:\n",
        "\n",
        "1) Utilizando um arquivo de dados com várias colunas (por exemplo, o arquivo ```dados_alunos.txt```), faça um histograma com os dados de cada uma das colunas. **Dica**: utilize o ***matplotlib*** para fazer os histogramas.\n",
        "\n",
        "2) Estude os métodos do módulo ```os``` e faça um script que liste todos os arquivos de um dado diretório assim como de seus subdiretórios. **Dica**: use o método ```walk```.\n",
        "\n",
        "3) **Reescreva o script dessa aula, ```lcount.py``` na forma de uma módulo. Qual o valor da variável ```__name__``` quando o módulo é importado?**\n",
        "\n",
        "4) Escreva uma função chamada ler_arquivos que tome como argumento um nome de arquivo, leia um arquivo com um número qualquer de colunas e retorne um **dicionário** que tenha como *keys* os números das colunas e como valor uma lista dos valores associados a cada coluna do arquivo. Use o arquivo dados_alunos.txt. Caso o arquivo tiver o cabeçalho das colunas, p.ex. dados_alunos_cabecalho.txt, a key deve ser a palavra do cabeçalho.\n",
        "\n",
        "   * Caso não queria usar dicionários, você pode pensar em uma estrutura de dados alternativa para armazenar os dados do arquivo, como por exemplo uma ou mais listas.\n",
        "   * Também pode usar funções já existentes nos módulos de Python `numpy` ou `pandas` para ler arquivos de texto e armazenar nas estruturas desses módulos (`DataFrames` ou `arrays`).\n",
        "   * Adapte a função acima para, se ocorrer um erro ao abrir, ler ou fechar arquivos, o programa capturar a exceção, imprimir uma mensagem de erro e sair.\n",
        "\n",
        "5) **Adapte o script acima para ser um módulo.**\n",
        "\n",
        "**6) Escreva um script que importe o módulo criado acima e faça um histograma para cada coluna do arquivo, com seus respectivos valores. Use o dicionário ou a estrutura de dados criada em 1).**\n",
        "\n",
        "7) Adicione uma função ao script criado em 6), que calcule o desvio padrão amostral, desvio padrão populacional e média de cada distribuição representada nos histogramas acima. Para tal, use o módulo `statistics` do Python ou outro módulo de Python (`numpy`, `scipy`)."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1IieW6gfO1ZJ"
      },
      "source": [
        "1."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "81-62YFUO0-D",
        "outputId": "1ee386fa-a49a-4270-b4b2-dd1c2e1583d3"
      },
      "source": [
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "df = pd.read_csv('dados_alunos_cabecalho.csv')\n",
        "\n",
        "with open('dados_alunos_cabecalho.txt', 'w') as f:\n",
        "  df.to_string(f)\n",
        "\n",
        "df.head()\n",
        "\n",
        "idade = df['I(a)']\n",
        "\n",
        "plt.figure(figsize = (8, 6))\n",
        "plt.hist(idade)\n",
        "plt.title('Distribuição de idades')\n",
        "plt.xlabel('Idade')\n",
        "plt.ylabel('Quantidade de alunos')\n",
        "plt.show()\n",
        "\n",
        "altura = df['A(m)']\n",
        "\n",
        "plt.figure(figsize = (8,6))\n",
        "plt.hist(altura)\n",
        "plt.title('Distribuição de altura')\n",
        "plt.xlabel('Altura')\n",
        "plt.ylabel('Alunos')\n",
        "plt.show()\n",
        "\n",
        "peso = df['M(Kg)']\n",
        "\n",
        "plt.figure(figsize = (8, 6))\n",
        "plt.hist(peso)\n",
        "plt.title('Distribuição de peso')\n",
        "plt.xlabel('Peso')\n",
        "plt.ylabel('Alunos')\n",
        "plt.show()\n",
        "\n"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}