麻省理工学院公开课：计算机科学及编程导论习题4下

习题3：

写一个函数postRetirement，参数：存款savings、投资账户（其实就是养老金么...）的年增长率growthRates是一个数组，数组的长度表示退休后的年份，也就是说这个数组的范围是从退休后的第一年到len（growthRates）年，当年消费expenses

假设账户结算利率在取存款前；

返回一个数组，包含每一年账户资金。

完全按下面来定义这个函数

def postRetirement(savings, growthRates, expenses):

第一年投资账户资金：savings * (1 + growthRates[0]) - expenses

第二年投资账户资金：( savings * (1 + growthRates[0]) - expenses) * (1 + growthRates[1]) - expense

第三年投资账户资金：(( savings * (1 + growthRates[0]) - expenses) * (1 + growthRates[1]) - expense) * (1 + growthRates[2])- expense

当第x年的时候账户资金： x-1年的savings * (1 + growthRate[x]) - expense

def postRetirement(savings, growthRates, expenses):

	def annual_fund_Retired(savings, growthRates, expenses):
		years = len(growthRates)
		if years > 1:
			years = years - 1
			annual_total = annual_fund_Retired(savings, growthRates[:years], expenses) * (1 + growthRates[years]) - expenses
			return annual_total
		elif years == 1:
			annual_total = savings * (1 + growthRates[0]) - expenses
			return annual_total
		else:
			return 0
	
	fund = []
	for time in range(0, len(growthRates)):
		fund.append(annual_fund_Retired(savings, growthRates[:time + 1], expenses))
	return fund

习题4：

写一个叫findMaxExpenses的函数，这个函数有五个参数：工资salary、工资投资百分比save、工作时投资账户的年增长率preRetireGrowthRates（数组）、退休后投资账户的年增长率postRetireGrowthRates（数组）、epsilon。

其实之前感觉不是很好很科学啊，因为退休后不拿的话不就代表挂了么...人能提前知道什么时候挂吗？...

使用二分法找出退休后到退休结束之间（不就是挂了么）每年固定能消费的数额，这个数额能保证退休结束时账户余额的绝对值小于epsilon(生不带来死不带去，一定要用光...），初始搜索的消费数额在0和退休开始的未取存款之间。

提示1#可以用习题2。

这个函数必须在每次迭代时候显示出当前预估的消费数额。

提示2#可以利用习题3

提示3#答案在0和初始未取存款+epsilon之间。

这样定义函数：

def findMaxExpenses(salary, save, preRetireGrowthRates, postRetireGrowthRates, epsilon):

第一年投资账户资金：savings * (1 + growthRates[0]) - expenses

第二年投资账户资金：( savings * (1 + growthRates[0]) - expenses) * (1 + growthRates[1]) - expense

第三年投资账户资金：(( savings * (1 + growthRates[0]) - expenses) * (1 + growthRates[1]) - expense) * (1 + growthRates[2])- expense

当第i年的时候账户资金： x-i年的savings * (1 + growthRates[x-i]) - expense

savings为常数，设为a；growthRate，设第一年的为b0、第二2为b1、第n年为bn-1；

expense为未知数，设为x；账户资金为y

第一年投资账户资金：y = -x + a + ab0

第二年投资账户资金：y = -(2 + b1)x + a + ab0 + ab1 + abob1

可以推导出资金账户的金额： y = -mx + n；m、n为正数

y的绝对值需要小于epsilon；

x在0和a+epsilon之间；

下图黄线为|y|。

low = 0，high = a + epsilon；guess = (low + high) / 2

当y>0时，如果y>epsilon，取右边，也就是 low = guess。

当y<0时，如果y>epsilon，取左边，也就是 high = guess,

def nestEggVariable(salary, save, growthRates):

	def annual_save(salary, save):
		annual_save = salary * save
		return annual_save
	
	def annual_fund_Variable(salary, save, growthRates):
		years = len(growthRates)
		if years > 0:
			years = years - 1
			annual_total = annual_fund_Variable(salary, save, growthRates[:years]) * (1 + growthRates[years]) + annual_save(salary, save)
			return annual_total
		else:
			return 0
	
	fund = []
	for time in range(0, len(growthRates)):
		fund.append(annual_fund_Variable(salary, save, growthRates[:time + 1]))
	return fund



def postRetirement(savings, growthRates, expenses):

	def annual_fund_Retired(savings, growthRates, expenses):
		years = len(growthRates)
		if years > 1:
			years = years - 1
			annual_total = annual_fund_Retired(savings, growthRates[:years], expenses) * (1 + growthRates[years]) - expenses
			return annual_total
		elif years == 1:
			annual_total = savings * (1 + growthRates[0]) - expenses
			return annual_total
		else:
			return 0
	
	fund = []
	for time in range(0, len(growthRates)):
		fund.append(annual_fund_Retired(savings, growthRates[:time + 1], expenses))
	return fund


def findMaxExpenses(salary, save, preRetireGrowthRates, postRetireGrowthRates, epsilon):

	savings = nestEggVariable(salary, save, preRetireGrowthRates).pop()
	high = savings
	low = 0
	count = 1

	while True:
		guess = (high + low) / 2
		retirement_fund = postRetirement(savings, postRetireGrowthRates, guess).pop()
		absolute = abs(retirement_fund)
		if absolute >= epsilon and count <= 100:
			print count,": estimated expense = ", guess
			if retirement_fund > 0:
				low = guess
			else:
				high = guess
			count = count + 1
		else: 
			break
	
	return guess

最后发上来几个验证的函数

def testNestEggFixed():
    salary     = 10000
    save       = 0.10
    growthRate = 0.15
    years      = 5
    savingsRecord = nestEggFixed(salary, save, growthRate, years)
    print savingsRecord
    # Output should have values close to:
    # [1000.0, 2150.0, 3472.5, 4993.375, 6742.3812499999995]

def testNestEggVariable():
    salary      = 10000
    save        = 0.10
    growthRates = [0.03, 0.04, 0.05, 0, 0.03]
    savingsRecord = nestEggVariable(salary, save, growthRates)
    print savingsRecord
    # Output should have values close to:
    # [1000.0, 2040.0, 3142.0, 4142.0, 5266.2600000000002]

def testPostRetirement():
    savings     = 100000
    growthRates = [0.10, 0.05, 0, 0.05, 0.01]
    expenses    = 30000
    savingsRecord = postRetirement(savings, growthRates, expenses)
    print savingsRecord
    # Output should have values close to:
    # [80000.000000000015, 54000.000000000015, 24000.000000000015,
    # -4799.9999999999854, -34847.999999999985]

def testFindMaxExpenses():
    salary                = 10000
    save                  = 0.10
    preRetireGrowthRates  = [0.03, 0.04, 0.05, 0, 0.03]
    postRetireGrowthRates = [0.10, 0.05, 0, 0.05, 0.01]
    epsilon               = .01
    expenses = findMaxExpenses(salary, save, preRetireGrowthRates,
                               postRetireGrowthRates, epsilon)
    print expenses
    # Output should have a value close to:
    # 1229.95548986