我正在做一些计算建模,其中我需要搜索间隔不规则且数据未排序的数据,所以这是场景
我的示例数据点
col1 col2
1 92
9 45
7 22
2 14
5 10
所以我正在寻找的搜索算法是这样的,
假设如果key = 2
那么函数应该返回2
的索引,因为它可用,例如我想搜索3
,它在col1
中不存在,因为它不可用,我需要在两个方向上搜索最近的值,即2
和5
的索引
如果是这样的话awk
确切的键可以使用这样的东西进行搜索
function search(Arr,key){
if((key in Arr))
return key
}
但我真的不知道,但在上下方向搜索最近的值,以防找不到确切的键
我希望我的要求是明确的,如果投反对票,也请发表您的评论,因为我的问题是因为声誉较低(本论坛新手),我无法投票获得有用的答案,请合作。
壳解决方案,
perl -lane'
BEGIN{ $k=pop }
push @r, [@F];
END {
for (sort{ $a->[0] <=> $b->[0] } @r) {
$v= $_->[0] <=> $k;
$h{$v} = $_->[1];
last if $v >0;
}
print join " ", ($h{0} or @h{-1,1});
}
' file 3
输出
14 10
因为没有排序,所以最好是同时搜索精确、下限和上限,只遍历一次数据。
#include <iostream>
#include <vector>
#include <utility>
using Samples = std::vector<std::pair<int,int>>;
Samples::const_iterator Find( Samples const & samp_, int val_, Samples::const_iterator & prev_, Samples::const_iterator & next_ ) {
auto end = std::end(samp_);
auto lower = end;
auto upper = end;
auto it = begin(samp_);
for( ; it!=end; ++it) {
if ( it->first == val_ )
return it;
if ( it->first < val_ && ( lower == end || lower->first < it->first ) )
lower = it;
else if ( it->first > val_ && ( upper == end || upper->first > it->first ) )
upper = it;
}
prev_ = lower;
next_ = upper;
return end;
}
std::ostream & operator<<( std::ostream & os, std::pair<int,int> const & p ) {
return os << "( " << p.first << ", " << p.second << " )";
}
int main() {
Samples samps { {1,92}, {9,45},{7,22},{2,14},{5,10} };
auto test = [&] ( int v ) {
Samples::const_iterator lower;
Samples::const_iterator upper;
auto result = Find( samps, v, lower, upper );
if ( result != end( samps ) ) {
std::cout << "found " << *result << std::endl;
} else {
std::cout << "not found ";
if ( lower != end( samps ) )
std::cout << "lower is " << *lower;
else
std::cout << "no lower";
if ( upper != end( samps ) )
std::cout << " upper is " << *upper;
else
std::cout << " no upper";
std::cout << std::endl;
}
};
test(2);
test(3);
test(12);
test(-1);
}
结果是:
found ( 2, 14 )
not found lower is ( 2, 14 ) upper is ( 5, 10 )
not found lower is ( 9, 45 ) no upper
not found no lower upper is ( 1, 92 )
在 Gnu Awk 版本 4 中,你可以使用如下PROCINFO["sorted_in"]
:
gawk -vkey=7 -f a.awk file
其中a.awk
是:
{
a[$1]=$2
}
END {
if (key in a)
print "Found key "key" with value "a[key]
else {
PROCINFO["sorted_in"]="@ind_num_asc"
for (i in a) {
if (i+0>key) { k=i; break}
j=i
}
if (j)
print "Prev key: "j
if(k)
print "Next key: "k
}
}
输出:
$gawk -vkey=6 -f a.awk file
Prev key: 5
Next key: 7
$gawk -vkey=5 -f a.awk file
Found key 5 with value 10