machinelearning_notebook/0_python/6_Function.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 函数"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在大部分时候，在一个算法中，需要重复执行一组语句，如果每次都重复写出来，不仅乏味而且编程效率比较低，降低程序的可读性。为了将重复的执行抽象出来，可使用函数将一组操作封装成一个整体，给一个名称和参数列表作为可变量的输入。\n",
    "\n",
    "Python中函数的定义为："
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```\n",
    "def funcname(arg1, arg2,... argN):\n",
    "    \n",
    "    ''' Document String'''\n",
    "\n",
    "    statements\n",
    "\n",
    "\n",
    "    return <value>\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "将上面的语法理解为，定义了一个名为`funcname`的函数，它接受`arg1,arg2，…argN`的参数。函数在执行语句后返回一个`<value>`。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hey Jack!\n",
      "Jack, How do you do?\n"
     ]
    }
   ],
   "source": [
    "print(\"Hey Jack!\")\n",
    "print(\"Jack, How do you do?\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "不需要每次都写上面的两个语句，可以通过定义一个函数来替换它，这个函数只需一行就能完成任务。\n",
    "\n",
    "定义一个函数 `first_func()`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def first_func():\n",
    "    print(\"Hey Jack!\")\n",
    "    print(\"Jack, How do you do?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hey Jack!\n",
      "Jack, How do you do?\n",
      "Hey Jack!\n",
      "Jack, How do you do?\n"
     ]
    }
   ],
   "source": [
    "first_func()\n",
    "funca=first_func\n",
    "funca()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**first_func()** 每一次只打印一个人的消息。我们可以让我们的函数 **first_func()** 接受参数，该参数将存储名称然后打印相应地接受字符串。为了这样做我们需要像所示的那样在函数内添加一个参数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def first_func(username):\n",
    "    print(\"Hey\", username + '!')\n",
    "    print(username + ',' ,\"How do you do?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Please enter your name : Jack\n"
     ]
    }
   ],
   "source": [
    "name1 = input('Please enter your name : ')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "名称“Willam”实际上存储在name1中。我们将这个变量传递给函数**first_func()** 作为变量username，因为它是为这个函数定义的变量。即name1作为用户名传递。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hey Jack!\n",
      "Jack, How do you do?\n"
     ]
    }
   ],
   "source": [
    "first_func(name1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "让我们通过定义另一个函数**second_func()** 来进一步简化它，该函数接受名称并将其存储在一个变量中，然后从函数本身内部调用**first_func()**。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def first_func(username):\n",
    "    print(\"Hey\", username + '!')\n",
    "    print(username + ',' ,\"How do you do?\")\n",
    "def second_func():\n",
    "    name = input(\"Please enter your name : \")\n",
    "    first_func(name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Please enter your name : Tom\n",
      "Hey Tom!\n",
      "Tom, How do you do?\n"
     ]
    }
   ],
   "source": [
    "second_func()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. 返回语句"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "当函数产生某个值，并且该值必须存储在一个变量中，或者需要返回或返回给主算法进行进一步操作时，使用return语句。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def times(x,y):\n",
    "    z = x*y\n",
    "    return z"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "上面定义的**times()** 函数接受两个参数并返回变量z，该变量包含两个参数的乘积结果。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20\n"
     ]
    }
   ],
   "source": [
    "c = times(4,5)\n",
    "print(c)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "z值存储在变量c中，可以用于进一步的操作。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "可以在return语句中使用整个语句本身，而不是声明另一个变量，如下所示。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def times(x,y):\n",
    "    '''This multiplies the two input arguments'''\n",
    "    return x*y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20\n"
     ]
    }
   ],
   "source": [
    "c = times(4,5)\n",
    "print(c)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "由于现在定义了**times()**，我们可以如上所示记录它。在**help()** 函数下调用**times()** 函数时返回该文档。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function times in module __main__:\n",
      "\n",
      "times(x, y)\n",
      "    This multiplies the two input arguments\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(times)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "times?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "也可以返回多个变量，但请记住顺序。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "eglist = [10,50,30,12,6,8,100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def egfunc(eglist):\n",
    "    highest = max(eglist)\n",
    "    lowest = min(eglist)\n",
    "    first = eglist[0]\n",
    "    last = eglist[-1]\n",
    "    return highest,lowest,first,last"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果调用函数时没有为其分配任何变量，则结果将在一个元组中返回。但是如果变量被提到了，那么结果会以特定的顺序分配给变量，这个顺序在return语句中声明。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100, 6, 10, 100)\n"
     ]
    }
   ],
   "source": [
    "a = egfunc(eglist)\n",
    "print(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " a = 100 \n",
      " b = 6 \n",
      " c = 10 \n",
      " d = 100\n"
     ]
    }
   ],
   "source": [
    "a,b,c,d = egfunc(eglist)\n",
    "print(' a =',a,'\\n b =',b,'\\n c =',c,'\\n d =',d)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. 隐式参数"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "当一个函数的参数在大多数情况下是常见的或者它是`隐式的`时，使用这个概念。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def implicit_add(x, addnumber=3):\n",
    "    return x+addnumber"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**implicit_add( )** 是一个函数接受两个参数，但大多数时候第一个参数只需要加3。因此，第二个参数被赋值为3。这里第二个参数是隐式的。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "现在，如果在调用**implicit_add()** 函数时没有定义第二个参数，则将其视为3。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "implicit_add(4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "但如果指定了第二个参数，则此值将覆盖分配给该参数的隐式值"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "implicit_add(4,4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "11"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "implicit_add(5, addnumber=6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. 任意数量的参数"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果函数要接受的参数数量未知，则在参数前使用星号。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_n(*args):\n",
    "    res = 0\n",
    "    reslist = []\n",
    "    for i in args:\n",
    "        reslist.append(i)\n",
    "    print(reslist)\n",
    "    return sum(reslist)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "上面的函数接受任意数量的参数，定义一个列表，并将所有参数附加到该列表中，并返回所有参数的总和。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 2, 3, 4, 5]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "15"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "add_n(1,2,3,4,5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 2, 3]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "add_n(1,2,3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "参数列表也可以通过 \"param_name = value\" 的形式传入到函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10, 20, 30]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "60"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def add_nd(**kwargs):\n",
    "    res = 0\n",
    "    reslist = []\n",
    "    for (k,v) in kwargs.items():\n",
    "        reslist.append(v)\n",
    "    print(reslist)\n",
    "    return sum(reslist)\n",
    "\n",
    "add_nd(x=10, y=20, c=30)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. 全局和局部变量"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在函数内部声明的变量是局部变量，生命周期限于函数执行期间；在函数外部声明的是全局变量。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "eg1 = [1,2,3,4,5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在下面的函数中，我们将在函数内部声明的列表中追加一个元素。函数内部声明的eg2变量是一个局部变量。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "def egfunc1():\n",
    "    eg1 = [1, 2, 3, 4, 5, 7]\n",
    "    print(\"egfunc1>> eg1: \", eg1)\n",
    "\n",
    "def egfunc2():\n",
    "    eg1.append(8)\n",
    "    print(\"egfunc2>> eg1: \", eg1)\n",
    "\n",
    "def egfunc3():\n",
    "    global eg1\n",
    "    eg1 = [5, 4, 3, 2, 1]\n",
    "    print(\"egfunc3>> eg1: \", eg1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "egfunc1>> eg1:  [1, 2, 3, 4, 5, 7]\n",
      "eg1:  [1, 2, 3, 4, 5] \n",
      "\n",
      "egfunc2>> eg1:  [1, 2, 3, 4, 5, 8]\n",
      "eg1:  [1, 2, 3, 4, 5, 8] \n",
      "\n",
      "egfunc3>> eg1:  [5, 4, 3, 2, 1]\n",
      "eg1:  [5, 4, 3, 2, 1] \n",
      "\n"
     ]
    }
   ],
   "source": [
    "egfunc1()\n",
    "print(\"eg1: \", eg1, \"\\n\")\n",
    "\n",
    "egfunc2()\n",
    "print(\"eg1: \", eg1, \"\\n\")\n",
    "\n",
    "egfunc3()\n",
    "print(\"eg1: \", eg1, \"\\n\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Lambda 函数"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "程序中有时需要临时使用一个简单的函数，单独定义出来比较费事，为了提高编程效率，Python等很多语言引入了`Lambda`函数，这些Lambda函数没有使用任何名称定义，只携带一个表达式，返回的是函数本身（类似函数指针或者函数对象）。Lambda函数在操作列表时非常方便。这些函数由关键字**lambda**定义，后面跟着变量、冒号和相应的表达式。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "z = lambda x: x * x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z(8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6, 8)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zz = lambda x, y: (x*y, x**y)\n",
    "zz(2, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "function"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(z)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "function"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def sqr(x):\n",
    "    return x*x\n",
    "\n",
    "type(sqr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.1 map"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**map( )** 函数主要执行分别为列表的每个元素定义的函数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "list1 = [1,2,3,4,5,6,7,8,9]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[3, 4, 5, 6, 7, 8, 9, 10, 11]\n"
     ]
    }
   ],
   "source": [
    "eg = map(lambda x:x+2, list1)\n",
    "print(list(eg))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 4, 9, 16, 25, 36, 49, 64, 81]\n"
     ]
    }
   ],
   "source": [
    "eg = map(sqr, list1)\n",
    "print(list(eg))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你也可以添加两个列表。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "list2 = [9,8,7,6,5,4,3,2,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10, 10, 10, 10, 10, 10, 10, 10, 10]\n"
     ]
    }
   ],
   "source": [
    "eg2 = map(lambda x,y:x+y, list1,list2)\n",
    "print(list(eg2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "不仅可以使用lambda函数，还可以使用其他内建函数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<map object at 0x7f7bb06e1610>\n"
     ]
    }
   ],
   "source": [
    "eg3 = map(str,eg2)\n",
    "print(eg3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.2 filter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**filter( )** 函数用于过滤列表中的值。 注意 **filter()** 函数返回一个新列表中的结果。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "list1 = [1,2,3,4,5,6,7,8,9]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "为了得到小于5的元素。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 2, 3, 4]\n"
     ]
    }
   ],
   "source": [
    "res = filter(lambda x:x<5,list1)\n",
    "print(list(res))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "注意当使用**map()** 时会发生什么"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<map at 0x7f7bb06c6590>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "map(lambda x:x<5, list1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们可以得出这样的结论，在**map()** 函数中返回的值为真，在使用**filter()** 函数时将返回特定的元素。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<filter at 0x7f7bb06ddf10>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filter(lambda x:x%4==0,list1)"
   ]
  }
 ],
 "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": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}