This can be done in O(n ) with O(1) space compexity . This problem can be
reduced to finding the kth smallest number in an unordered collection .
There exists an O(n) algo for this . For finding the 10th most frequent
number we should find  (n-10)th smallest number . Once found by doing an
linear scan we can finf max 10 elements . If we want to preserve the order
of 10 max elements we can sort it constant time .


-Thanks & Regards
K.Abhinav Raghunandan
Department of Computer Science Engg,
IIT Madras



On Mon, Dec 27, 2010 at 10:01 PM, DIPANKAR DUTTA <[email protected]
> wrote:

> hey...
> what's about the storage managment ?
> the search text may not be a fixed length string..
> is implementation of a node is fessible?
>
> one alternative solution may be associate a hash table along with this. but
> again the problem with collision ,, to avoid collision u can use perfect
> hashing with universal hash function.
>
> you can also use an alternative solution hash function and priority queue
> using MAX heap...
>
>
> On Mon, Dec 27, 2010 at 2:20 AM, Chi <[email protected]> wrote:
>
>>  A trie is a binary tree with it nodes has as many leafs as is suitable.
>> A binary tree has only 2 leafs per nodes. In a trie it happens that a node
>> doesn't contain any leafs. This happens over many nodes. This is
>> considerable a waste of space. A radix-trie tries to eliminate this empty
>> nodes by compressing them into 1 node. The difference between a radix-trie
>> and suffix-trie is that a suffix-trie can solve text-based problems. A
>> radix-trie is good to store a dictionnary and search for words. Or
>> associative arrays. Or ip addresses.
>>
>> Unlike balanced trees, radix trees permit lookup, insertion, and deletion
>> in O(k) time rather than O(log n). This doesn't seem like an advantage,
>> since normally k ≥ log n, but in a balanced tree every comparison is a
>> string comparison requiring O(k) worst-case time, many of which are slow in
>> practice due to long common prefixes. In a trie, all comparisons require
>> constant time, but it takes m comparisons to look up a string of length m.
>> Radix trees can perform these operations with fewer comparisons and require
>> many fewer nodes.
>> K=key
>> n=string
>>
>> I have a written a kart-trie in php:
>> https://sourceforge.net/projects/kart-trie
>>
>> The only difference to a radix-trie is that the decision to branch either
>> left or right is used with a key of 32-bit.
>>
>> ----- Ursprüngliche Mitteilung -----
>> > @Chi: Would you please explain the process in a little bit more
>> > details? It will be helpful.
>> > Is Trie and Radix-Trie same?
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups "Algorithm Geeks" group. To post to this group, send email to
>> > [email protected]. To unsubscribe from this group, send email
>> > to [email protected]. For more options, visit this
>>
>> > group at http://groups.google.com/group/algogeeks?hl=en.
>> >
>>
>>  --
>> You received this message because you are subscribed to the Google Groups
>> "Algorithm Geeks" group.
>> To post to this group, send email to [email protected].
>> To unsubscribe from this group, send email to
>> [email protected]<algogeeks%[email protected]>
>> .
>> For more options, visit this group at
>> http://groups.google.com/group/algogeeks?hl=en.
>>
>
>
>
> --
> DIPANKAR DUTTA
> M-TECH,Computer Science & Engg.
> E&C Dept,IIT ROORKEE
> Uttarakhand , India – 247667
> -------------------------------------------
> website:http://people.iitr.ernet.in/shp/09535009/Website/index.html
> ph no-09045809987
> Lab: 286454
> email:[email protected] <email%[email protected]>
>
>  --
> You received this message because you are subscribed to the Google Groups
> "Algorithm Geeks" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to
> [email protected]<algogeeks%[email protected]>
> .
> For more options, visit this group at
> http://groups.google.com/group/algogeeks?hl=en.
>

-- 
You received this message because you are subscribed to the Google Groups 
"Algorithm Geeks" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/algogeeks?hl=en.

Reply via email to