假设我有一个包含3个数字的数组:
NSArray *array = @[@1, @2, @3];
我想做所有组合而不重复
所以我需要的是这个:
(1)
(2)
(3)
(1,2)
(2,3)
(1,3)
(1,2,3)
我目前的代码是:
NSArray *array = @[@1, @2, @3];
int numberOfCardsOTable = [array count];
//NSLog(@"array = %@", array);
for (int lenghtOfArray = 1; lenghtOfArray <= numberOfCardsOTable; lenghtOfArray++)
{
for (int i = 0; i < numberOfCardsOTable; i++)
{
// array bound check
if (i + lenghtOfArray > numberOfCardsOTable) {
continue;
}
NSArray *subArray = [[NSMutableArray alloc] init];
subArray = [array subarrayWithRange:NSMakeRange(i, lenghtOfArray)];
NSLog(@"array = %@", subArray);
}
}
但缺少此代码(1,3)。
我需要为最多8个数字的源数组执行此操作
有8个数字,有255种组合,我的算法会遗漏很多,因此会有很多if
s。
答案 0 :(得分:6)
由于您似乎希望组合与原始组合的顺序相同,所以您所做的与计数到2 num_choices 并选择与设置位对应的对象相同。你可以通过我为NSIndexSet
编写的类别方法提供一些帮助,使这个变得非常简单。
@implementation NSIndexSet (WSSNoncontiguous)
+ (instancetype)WSSIndexSetFromMask:(uint64_t)mask
{
NSMutableIndexSet * set = [NSMutableIndexSet indexSet];
for( uint64_t i = 0; i < 64; i++ ){
if( mask & (1ull << i) ){
[set addIndex:i];
}
}
return set;
}
@end
这将创建一个NSIndexSet
,其内容是掩码中设置的位的索引。然后,您可以使用-[NSArray objectsAtIndexes:]
的索引集来获取组合:
NSArray * choices = @[...];
uint64_t num_combos = 1ull << [choices count]; // 2**count
NSMutableArray * combos = [NSMutableArray new];
for( uint64_t i = 1; i < num_combos; i++ ){
NSIndexSet * indexes = [NSIndexSet WSSIndexSetFromMask:i];
[combos addObject:[choices objectsAtIndexes:indexes]];
}
显然,这仅适用于成员数量为64或更少的choices
,但无论如何最终会成为非常大量的组合。
答案 1 :(得分:1)
范围永远不会适用于这种情况。
有趣的是,您可能需要考虑以下事项。
NSArray *array = @[@1, @2, @3];
//NSArray *array = @[@1, @2, @3, @4];
int numberOfCardsOTable = [array count];
//this can be calculated too - a row in Pascal's triangle
NSArray *pascalsRow = @[@1, @3, @3, @1];
//NSArray *pascalsRow = @[@1, @4, @6, @4, @1];
int pIndex = 1;
int endIdx = [[pascalsRow objectAtIndex:pIndex] integerValue];
int outputLength;
//process the number of expected terms in pascal row
for (int i = 0; i < [pascalsRow count]; i++)
{
//skipping first term
outputLength = i;
if(outputLength > 0)
{
for (int j = i; j <= endIdx; j++)
{
if(outputLength > 1)
{
for(int k = 1; k <= endIdx; k++)
{
NSLog(@"j = %i, k = %i, ... outputLength = %i", j, k, outputLength);
}
j = endIdx;
}
else
NSLog(@"j = %i, ... outputLength = %i", j, outputLength);
}
if(pIndex < numberOfCardsOTable)
{
pIndex++;
NSLog(@"pIndex = %i, endIdx = %i", pIndex, endIdx);
endIdx = [[pascalsRow objectAtIndex:pIndex] integerValue];
}
if(endIdx == 1 && outputLength == numberOfCardsOTable)
NSLog(@"... outputLength = %i", outputLength);
}
}
我留下p = 4注释掉,以防你想快速查看该用例。这种方法至少可以为您提供正确数量的术语和长度。我会留给你构建所需的输出。
输出:
[32771:70b] j = 1, ... outputLength = 1
[32771:70b] j = 2, ... outputLength = 1
[32771:70b] j = 3, ... outputLength = 1
[32771:70b] pIndex = 2, endIdx = 3
[32771:70b] j = 2, k = 1, ... outputLength = 2
[32771:70b] j = 2, k = 2, ... outputLength = 2
[32771:70b] j = 2, k = 3, ... outputLength = 2
[32771:70b] pIndex = 3, endIdx = 3
[32771:70b] ... outputLength = 3