Mirror/zeek
1
0
Fork 0
mirror of https://github.com/zeek/zeek.git synced 2025-10-04 23:58:20 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

update hll documentation, make a few functions private and create

Browse source 
a new copy constructor.
This commit is contained in:
Bernhard Amann 2013-09-16 10:40:25 -07:00
parent eb1d7ccc4a
commit c0f780c728
3 changed files with 108 additions and 63 deletions
Download patch file Download diff file Expand all files Collapse all files

29
src/probabilistic/CardinalityCounter.cc
View file

@ -15,6 +15,9 @@ int CardinalityCounter::OptimalB(double error, double confidence)
double initial_estimate = 2 * (log(1.04) - log(error)) / log(2);
int answer = (int) floor(initial_estimate);
// k is the number of standard deviations that we have to go to have
// a confidence level of conf.
double k = 0;
do {
@ -54,6 +57,12 @@ void CardinalityCounter::Init(uint64 size)
V = m;
}
CardinalityCounter::CardinalityCounter(CardinalityCounter& other)
{
Init(other.GetM());
Merge(&other);
}
CardinalityCounter::CardinalityCounter(double error_margin, double confidence)
{
int b = OptimalB(error_margin, confidence);
@ -107,7 +116,16 @@ void CardinalityCounter::AddElement(uint64 hash)
buckets[index] = temp;
}
double CardinalityCounter::Size()
/**
* Estimate the size by using the the "raw" HyperLogLog estimate. Then,
* check if it's too "large" or "small" because the raw estimate doesn't
* do well in those cases.
* Thus, we correct for those errors as specified in the paper.
*
* Note - we deviate from the HLL algorithm in the paper here, because
* of our 64-bit hashes.
**/
double CardinalityCounter::Size() const
{
double answer = 0;
for ( unsigned int i = 0; i < m; i++ )
@ -126,8 +144,11 @@ double CardinalityCounter::Size()
return -pow(2, 64) * log(1 - (answer / pow(2, 64)));
}
void CardinalityCounter::Merge(CardinalityCounter* c)
bool CardinalityCounter::Merge(CardinalityCounter* c)
{
if ( m != c->GetM() )
return false;
uint8_t* temp = c->GetBuckets();
V = 0;
@ -140,6 +161,8 @@ void CardinalityCounter::Merge(CardinalityCounter* c)
if ( buckets[i] == 0 )
++V;
}
return true;
}
uint8_t* CardinalityCounter::GetBuckets()
@ -147,7 +170,7 @@ uint8_t* CardinalityCounter::GetBuckets()
return buckets;
}
uint64 CardinalityCounter::GetM()
uint64 CardinalityCounter::GetM() const
{
return m;
}