统计算法_探索性统计

import random case_list = [] #准备个容器，放样例 #随机生成10个1到30的小数，不能重复，把10个小数放到容器里 while len(case_list) < 10: rand_float = random.uniform(1,30) if rand_float in case_list: continue case_list.append(rand_float) case_list = [round(case,2) for case in case_list] #格式化一下，不然太长不好看，当然这个因需要而定，我为了显示好看，所以格式化保留两位小数 #我先执行了个结果 print case_list #结果是，[5.77, 27.21, 27.12, 12.65, 23.12, 25.56, 18.6, 23.75, 4.1, 25.13] case_list[6] = 23.12 #为了后面众数，弄个相同值出来 print str(case_list) #单独打下面了，方便看 '[5.77, 27.21, 27.12, 12.65, 23.12, 25.56, 23.12, 23.75, 4.1, 25.13]' case_list.sort() #排序,这个算法有很多,单独写,不在这里啰嗦 #先写个累加函数，为了方便，就不try了，默认输入的结果都是对的 def sum_fun(xlist): n = 0 for x in xlist: n += x return n #统计数据量函数 def len_fun(xlist): n = 0 for x in xlist: n += 1 return n #累乘，和累加一样 def multiply_fun(xlist): n = 1 for x in xlist: n *= x return n #1、算数平均数：加和/个数 def sum_mean_fun(case_list):
  sum_mean_num = sum_fun(case_list)/len_fun(case_list)
  return sum_mean_num #结果是19.753 #2、算数平均数回报,平均回报率计算, def sum_mean_rate(case_list):
  '((case_list[1]-case_list[0])/case_list(0)+(case_list[2]-case_list[1])/case_list(1)+...(case_list[n]-case_list[n-1])/case_list(n-1))/len(case_list-1)'
  n = 1
  rate_return = [] #存放回报率
  while n < len_fun(case_list):
    rate = (case_list[n] - case_list[n-1])/case_list[n-1]
    rate_return.append(rate)
    n += 1
  mean_rate_num = sum_fun(rate_return)/len_fun(rate_return)
  return mean_rate_num #3、中位数,list中间的数,如果count是基数index = len(list)/2,如果是偶数index1 = (len(list)/2 index2 = (len(list)/2)+1)  def median_fun(case_list):
  if len_fun(case_list)%2 == 1:
    median = case_list[(len_fun(case_list)+1)/2-1]
  else:
    median = (case_list[len_fun(case_list)/2-1] + case_list[len_fun(case_list)/2])/2
  return median #4、众数,存在最多的数 def modes_fun(case_list):
  case_list_delre = list(set(case_list)) #去重,这个如果自己写的话可以用分治法,有兴趣的话可以自己写个
  count_max = 0
  for case_part in case_list_delre:
    case_count = case_list.count(case_part)
    if case_count > count_max:
      count_max = case_count
      max_return = case_part
  if count_max == 1:
    return None
  mode = max_return
  return mode #5、极差,最大-最小,因为已经拍好序,所以index min - index max def ext_minus_fun(case_list):
  ext_minus_num = case_list[len_fun(case_list)-1] - case_list[0]
  return ext_minus_num #6、四分位数,箱图用的,可以避免极值的影响,分别是index1 = len(list)/4 index2 = 3*len(list)/4 def four_bit_fun(case_list):
  Q1 = case_list[len_fun(case_list)/4]
  Q2 = case_list[3*len_fun(case_list)/4]
  return Q1,Q2 #7、几何平均数,和算数平均数不同,把所有都乘过^(1/len(list)) def geom_mean_fun(case_list):
  geom_mean_num = multiply_fun(case_list) (1.0/len_fun(case_list))
  return geom_mean_num #8、几何平均回报,R = 回报,((1+R1)(1+R2)...(1+Rn))^1/(len(count(R))-1) def geom_mean_rate(case_list):
  n = 1
  rate_return = [] #存放回报率
  while n < len_fun(case_list):
    rate = (case_list[n] - case_list[n-1])/case_list[n-1]
    rate_return.append(rate)
    n += 1
  rate_return = [1+rate for rate in rate_return]
  geom_mean_rate_num = multiply_fun(rate_return) (1.0/len_fun(rate_return)) - 1
  return geom_mean_rate_num

# 验证
if __name__ == '__main__':
  rand_list = create_rand_list(1,30,10)
  rand_list.sort()
  print rand_list
  sum_mean_num = sum_mean_fun(rand_list)
  print sum_mean_num
  mean_rate_num = sum_mean_rate(rand_list)
  print mean_rate_num
  median_num = median_fun(rand_list)
  print median_num
  modes_num = modes_fun(rand_list)
  print modes_num
  ext_minus_num = ext_minus_fun(rand_list)
  print ext_minus_num
  four_bit_q1,four_bit_q2 = four_bit_fun(rand_list)
  print four_bit_q1
  print four_bit_q2
  geom_mean_num = geom_mean_fun(rand_list)
  print geom_mean_num
  geom_mean_rate_num = geom_mean_rate(rand_list)
  print geom_mean_rate_num