简单的 bruce force, 算法如下, 未能证明其正确性.
给定一组数据 (1,2,3,4) 和 操作符 (-,+,*)
则有两种可能结果
((1-2) + 3) * 4
((1-2) * (3+4))
再通过排列 给定的数据和操作符 来重复上述操作
请大家指点指点
- 代码语法等方面如何改进
- 算法上是否正确、如何改进
- 其他
谢谢!
给定一组数据 (1,2,3,4) 和 操作符 (-,+,*)
则有两种可能结果
((1-2) + 3) * 4
((1-2) * (3+4))
再通过排列 给定的数据和操作符 来重复上述操作
请大家指点指点
- 代码语法等方面如何改进
- 算法上是否正确、如何改进
- 其他
谢谢!
#!/usr/bin/python
from __future__ import division
import os
# cant use: y == 0 and 0 or x / y
# because 0 eval to false, thus when y==0, it still performs x/y
def div (x,y):
if y==0: return 0
return x / y
operation = {'+' : lambda x,y : x + y,
'-' : lambda x,y : x - y,
'*' : lambda x,y : x * y,
'/' : lambda x,y : div(x,y)}
all_ops = operation.keys()
all_number_cand = (1,2,3,4,5,6,7,8,9,10)
def op_perm (ops=all_ops):
return [(i,j,k) for i in ops for j in ops for k in ops]
def number_perm (numbers=all_number_cand):
return [(i,j,k,l) for i in numbers for j in numbers for k in numbers for l in numbers]
### TODO increase the performace, it takes 10s
def main ():
result = []
for n_tuple in number_perm():
for o_tuple in op_perm():
o_stack = list(o_tuple)
n_stack = list(n_tuple)
n_queue = list(n_tuple)
while o_stack:
op = operation[o_stack.pop()]
n_stack.append(op(n_stack.pop(), n_stack.pop()))
n_queue.insert(0, op(n_queue.pop(), n_queue.pop()))
if 24 in n_stack or 24 in n_queue:
result.append((' ').join([str(n_tuple),str(o_tuple),str(n_stack),str(n_queue)]))
file = open('cal_res.txt','a')
for item in result:
file.writelines(item + os.linesep)
file.close()
if __name__=='__main__':
open('cal_res.txt','wb').write('') # any quick method?
main()