将连续的整数组合在一起

时间:2018-04-12 04:09:37

标签: python algorithm list

拥有以下代码:

import sys


ints = [1,2,3,4,5,6,8,9,10,11,14,34,14,35,16,18,39,10,29,30,14,26,64,27,48,65]
ints.sort()
ints = list(set(ints))

c = {}

for i,v in enumerate(ints):

    if i+1 >= len(ints):
        continue

    if ints[i+1] == v + 1 or ints[i-1] == v - 1:

        if len(c) == 0:
            c[v] = [v]
            c[v].append(ints[i+1])
        else:
            added=False
            for x,e in c.items():
                last = e[-1]
                if v in e:
                    added=True
                    break

                if v - last == 1:
                    c[x].append(v)
                    added=True

            if added==False:
                c[v] = [v]
    else:
        if v not in c:
            c[v] = [v]



print('input ', ints)
print('output ', c))

目标:

给定一个整数列表,创建一个包含连续整数的字典,以减少列表的总长度。

以下是我当前解决方案的输出:

input  [1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 14, 16, 18, 26, 27, 29, 30, 34, 35, 39, 48, 64, 65]
output  {1: [1, 2, 3, 4, 5, 6], 8: [8, 9, 10, 11], 14: [14], 16: [16], 18: [18], 26: [26, 27], 29: [29, 30], 34: [34, 35], 39: [39], 48: [48], 64: [64]}

条件/约束:

  • 如果当前整数是a)在现有列表中或b)是现有列表中的最后一项,我们不想为此项创建另一个列表。 即在1-5范围内,当我们到达3时,不要创建列表3,4,而是将3附加到现有列表[1,2] < / LI>

我当前的迭代工作正常,但由于for x,e in c.items()现有列表检查,列表越大,指数越小。

如何在实现相同结果的同时加快速度?

新解决方案(使用19,000个整数的输入列表从13秒到0.03秒):

c = {}

i = 0

last_list = None

while i < len(ints):
    cur = ints[i]

    if last_list is None:
        c[cur] = [cur]
        last_list = c[cur]

    else:

        if last_list[-1] == cur-1:
            last_list.append(cur)
        else:
            c[cur] = [cur]
            last_list = c[cur]

    i += 1

6 个答案:

答案 0 :(得分:3)

由于您有连续数字列表,我建议您使用range个对象而不是list s:

d, head = {}, None
for x in l:
    if head is None or x != d[head].stop:
        head = x
    d[head] = range(head, x+1)

答案 1 :(得分:2)

如果您使用@Override protected void onCreate(Bundle savedInstanceState) { super.onCreate(savedInstanceState); setContentView(R.layout.activity_add_question); add_question = findViewById(R.id.addquestion); question_EditText = findViewById(R.id.question_add); add_question.setOnClickListener(new View.OnClickListener() { @Override public void onClick(View v) { final String question = question_EditText.getText().toString(); Log.d("finds","question1:"+question); if (question.length()==0) { new AlertDialog.Builder(AddQuestionActivity.this) .setTitle(R.string.add_error) .setMessage(R.string.add_blank) .setPositiveButton(R.string.ok, null) .setNeutralButton(R.string.cancel, null) .show(); } } }); } 循环并且只是跟踪当前列表,那么解决方案很简单。当你找到差距时,不要忘记制作新的清单:

for

答案 2 :(得分:1)

有一个很棒的名为more_itertools的库,它有一个名为consecutive_groups()的方法:

import more_itertools as mit
x = [1,2,3,4,5,6,8,9,10,11,14,34,14,35,16,18,39,10,29,30,14,26,64,27,48,65]
x = [list(j) for j in mit.consecutive_groups(sorted(list(set(x))))]
# [[1, 2, 3, 4, 5, 6], [8, 9, 10, 11], [14], [16], [18], [26, 27], [29, 30], [34, 35], [39], [48], [64, 65]]
dct_x = {i[0]: i for i in x}
print(dct_x)

输出:

{1: [1, 2, 3, 4, 5, 6], 8: [8, 9, 10, 11], 14: [14], 16: [16], 18: [18], 26: [26, 27], 29: [29, 30], 34: [34, 35], 39: [39], 48: [48], 64: [64, 65]}

还有一条评论,您希望在转换为集合之后进行排序,因为集合是无序的。

答案 3 :(得分:0)

可以在O(n)(线性)复杂性中解决此任务。保持简单:

integers = [1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 14, 16, 18, 26, 27, 29, 30, 34, 35, 39, 48, 64, 65]

helper = []
counter = 0
while counter < len(integers):
  if not helper or helper[-1] + 1 != integers[counter]:
    print('gap found', integers[counter])  # do your logic
  helper.append(integers[counter])
  counter += 1

上述算法假定输入列表已经排序。它给了我们巨大的优势。同时,在运行此算法之前,可以明确地对整数列表进行排序。然后,解决方案的总复杂性将是:O(n * log n) + O(n),这是有效的O(n * log n)O(n * log n)是排序过程的复杂性。

我建议记住这个非常有用的技巧,在接近任务以便将来使用之前使用排序。

答案 4 :(得分:0)

这是一个简单的实现,使用列表切片实现您的目标:

integers = [1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 14, 16, 18, 26, 27, 29, 30, 34, 35, 39, 48, 64, 65]
for i, integer in enumerate(integers):
    if i == 0:
        out_dict = {}
        start = 0
    else:
        if integer != prev_integer + 1:
            out_dict[integers[start]] = integers[start:i]
            start = i
        if i == len(integers) - 1:
            out_dict[integers[start]] = integers[start:]
    prev_integer = integer

>>>out_dict =  {1: [1, 2, 3, 4, 5, 6], 8: [8, 9, 10, 11], 14: [14], 16: [16], 18: [18], 26: [26, 27], 29: [29, 30], 34: [34, 35], 39: [39], 48: [48], 64: [64]}

注意:字典可能不会按升序排序,因为dict类型没有排序。

答案 5 :(得分:0)

您可以尝试使用itertools,但我想尝试递归:

input_dta=[1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 14, 16, 18, 26, 27, 29, 30, 34, 35, 39, 48, 64, 65]

final_=[]

def consecutives(data):
    sub_final=[]

    if not data:
        return 0
    else:
        for i,j in enumerate(data):
            try:

                if abs(data[i]-data[i+1])==1:
                    sub_final.extend([data[i],data[i+1]])



                else:
                    if sub_final:
                        final_.append(set(sub_final))




                    return consecutives(data[i+1:])
            except IndexError:
                pass

    final_.append(set(sub_final))


consecutives(input_dta)



print(final_)

输出:

[{1, 2, 3, 4, 5, 6}, {8, 9, 10, 11}, {26, 27}, {29, 30}, {34, 35}, {64, 65}]