// I embarassingly don't know how templating works.
void QuoteGraph::CompareUsingMinHash(TVec<THash<TMd5Sig, TIntSet> >& BucketsVector) {
THashSet<TIntPr> EdgeCache;
int Count = 0;
int RealCount = 0;
Err("Beginning edge creation step...\n");
for (int i = 0; i < BucketsVector.Len(); i++) {
Err("Processing band signature %d of %d - %d signatures\n", i+1, BucketsVector.Len(), BucketsVector[i].Len());
TVec<TMd5Sig> Buckets;
BucketsVector[i].GetKeyV(Buckets);
TVec<TMd5Sig>::TIter BucketEnd = Buckets.EndI();
for (TVec<TMd5Sig>::TIter BucketSig = Buckets.BegI(); BucketSig < BucketEnd; BucketSig++) {
TIntSet Bucket = BucketsVector[i].GetDat(*BucketSig);
Count += Bucket.Len() * (Bucket.Len() - 1) / 2;
for (TIntSet::TIter Quote1 = Bucket.BegI(); Quote1 < Bucket.EndI(); Quote1++) {
TIntSet::TIter Quote1Copy = Quote1;
Quote1Copy++;
for (TIntSet::TIter Quote2 = Quote1Copy; Quote2 < Bucket.EndI(); Quote2++) {
if (!EdgeCache.IsKey(TIntPr(Quote1.GetKey(), Quote2.GetKey())) && !EdgeCache.IsKey(TIntPr(Quote2.GetKey(), Quote1.GetKey()))) {
EdgeCache.AddKey(TIntPr(Quote1.GetKey(), Quote2.GetKey()));
EdgeCache.AddKey(TIntPr(Quote2.GetKey(), Quote1.GetKey()));
RealCount++;
AddEdgeIfSimilar(Quote1.GetKey(), Quote2.GetKey());
}
}
}
}
}
fprintf(stderr, "NUMBER OF COMPARES: %d\n", Count);
fprintf(stderr, "NUMBER OF REAL COMPARES: %d\n", RealCount);
}
void QuoteGraph::CompareUsingShingles(THash<TMd5Sig, TIntSet>& Shingles) {
int Count = 0;
int RealCount = 0;
TVec<TMd5Sig> ShingleKeys;
Shingles.GetKeyV(ShingleKeys);
THashSet<TIntPr> EdgeCache;
for (int i = 0; i < ShingleKeys.Len(); i++) {
if (i % 100 == 0) {
Err("Processed %d out of %d shingles, count = %d\n", i, ShingleKeys.Len(), Count);
}
TIntSet Bucket;
Shingles.IsKeyGetDat(ShingleKeys[i], Bucket);
for (TIntSet::TIter Quote1 = Bucket.BegI(); Quote1 < Bucket.EndI(); Quote1++) {
TIntSet::TIter Quote1Copy = Quote1;
Quote1Copy++;
for (TIntSet::TIter Quote2 = Quote1Copy; Quote2 < Bucket.EndI(); Quote2++) {
if (!EdgeCache.IsKey(TIntPr(Quote1.GetKey(), Quote2.GetKey())) && !EdgeCache.IsKey(TIntPr(Quote2.GetKey(), Quote1.GetKey()))) {
EdgeCache.AddKey(TIntPr(Quote1.GetKey(), Quote2.GetKey()));
EdgeCache.AddKey(TIntPr(Quote2.GetKey(), Quote1.GetKey()));
RealCount++;
AddEdgeIfSimilar(Quote1.GetKey(), Quote2.GetKey());
}
}
}
int Len = Bucket.Len() * (Bucket.Len() - 1) / 2;
Count += Len;
}
fprintf(stderr, "NUMBER OF COMPARES: %d\n", Count);
fprintf(stderr, "NUMBER OF REAL COMPARES: %d\n", RealCount);
}
double TAGMFit::SelectLambdaSum(const TFltV& NewLambdaV, const TIntSet& ComK) {
double Result = 0.0;
for (TIntSet::TIter SI = ComK.BegI(); SI < ComK.EndI(); SI++) {
IAssert(NewLambdaV[SI.GetKey()] >= 0);
Result += NewLambdaV[SI.GetKey()];
}
return Result;
}
// YES I COPIED AND PASTED CODE my section leader would be so ashamed :D
void LSH::MinHash(THash<TMd5Sig, TIntSet>& ShingleToQuoteIds,
TVec<THash<TIntV, TIntSet> >& SignatureBandBuckets) {
TRnd RandomGenerator; // TODO: make this "more random" by incorporating time
for (int i = 0; i < NumBands; ++i) {
THash < TInt, TIntV > Inverted; // (QuoteID, QuoteSignatureForBand)
THash < TIntV, TIntSet > BandBuckets; // (BandSignature, QuoteIDs)
for (int j = 0; j < BandSize; ++j) {
// Create new signature
TVec < TMd5Sig > Signature;
ShingleToQuoteIds.GetKeyV(Signature);
Signature.Shuffle(RandomGenerator);
// Place in bucket - not very efficient
int SigLen = Signature.Len();
for (int k = 0; k < SigLen; ++k) {
TIntSet CurSet = ShingleToQuoteIds.GetDat(Signature[k]);
for (TIntSet::TIter l = CurSet.BegI(); l < CurSet.EndI(); l++) {
TInt Key = l.GetKey();
if (Inverted.IsKey(Key)) {
TIntV CurSignature = Inverted.GetDat(Key);
if (CurSignature.Len() <= j) {
CurSignature.Add(k);
Inverted.AddDat(Key, CurSignature);
}
} else {
TIntV NewSignature;
NewSignature.Add(k);
Inverted.AddDat(Key, NewSignature);
}
}
}
}
TIntV InvertedKeys;
Inverted.GetKeyV(InvertedKeys);
TInt InvertedLen = InvertedKeys.Len();
for (int k = 0; k < InvertedLen; ++k) {
TIntSet Bucket;
TIntV Signature = Inverted.GetDat(InvertedKeys[k]);
if (BandBuckets.IsKey(Signature)) {
Bucket = BandBuckets.GetDat(Signature);
}
Bucket.AddKey(InvertedKeys[k]);
BandBuckets.AddDat(Signature, Bucket);
}
SignatureBandBuckets.Add(BandBuckets);
Err("%d out of %d band signatures computed\n", i + 1, NumBands);
}
Err("Minhash step complete!\n");
}
/// Newton method: DEPRECATED
int TAGMFast::MLENewton(const double& Thres, const int& MaxIter, const TStr PlotNm) {
TExeTm ExeTm;
int iter = 0, PrevIter = 0;
TIntFltPrV IterLV;
double PrevL = TFlt::Mn, CurL;
TUNGraph::TNodeI UI;
TIntV NIdxV;
G->GetNIdV(NIdxV);
int CID, UID, NewtonIter;
double Fuc, PrevFuc, Grad, H;
while(iter < MaxIter) {
NIdxV.Shuffle(Rnd);
for (int ui = 0; ui < F.Len(); ui++, iter++) {
if (! PlotNm.Empty() && iter % G->GetNodes() == 0) {
IterLV.Add(TIntFltPr(iter, Likelihood(false)));
}
UID = NIdxV[ui];
//find set of candidate c (we only need to consider c to which a neighbor of u belongs to)
TIntSet CIDSet;
UI = G->GetNI(UID);
if (UI.GetDeg() == 0) { //if the node is isolated, clear its membership and skip
if (! F[UID].Empty()) { F[UID].Clr(); }
continue;
}
for (int e = 0; e < UI.GetDeg(); e++) {
if (HOVIDSV[UID].IsKey(UI.GetNbrNId(e))) { continue; }
TIntFltH& NbhCIDH = F[UI.GetNbrNId(e)];
for (TIntFltH::TIter CI = NbhCIDH.BegI(); CI < NbhCIDH.EndI(); CI++) {
CIDSet.AddKey(CI.GetKey());
}
}
for (TIntFltH::TIter CI = F[UID].BegI(); CI < F[UID].EndI(); CI++) { //remove the community membership which U does not share with its neighbors
if (! CIDSet.IsKey(CI.GetKey())) {
DelCom(UID, CI.GetKey());
}
}
if (CIDSet.Empty()) { continue; }
for (TIntSet::TIter CI = CIDSet.BegI(); CI < CIDSet.EndI(); CI++) {
CID = CI.GetKey();
//optimize for UID, CID
//compute constants
TFltV AlphaKV(UI.GetDeg());
for (int e = 0; e < UI.GetDeg(); e++) {
if (HOVIDSV[UID].IsKey(UI.GetNbrNId(e))) { continue; }
AlphaKV[e] = (1 - PNoCom) * exp(- DotProduct(UID, UI.GetNbrNId(e)) + GetCom(UI.GetNbrNId(e), CID) * GetCom(UID, CID));
IAssertR(AlphaKV[e] <= 1.0, TStr::Fmt("AlphaKV=%f, %f, %f", AlphaKV[e].Val, PNoCom.Val, GetCom(UI.GetNbrNId(e), CID)));
}
Fuc = GetCom(UID, CID);
PrevFuc = Fuc;
Grad = GradientForOneVar(AlphaKV, UID, CID, Fuc), H = 0.0;
if (Grad <= 1e-3 && Grad >= -0.1) { continue; }
NewtonIter = 0;
while (NewtonIter++ < 10) {
Grad = GradientForOneVar(AlphaKV, UID, CID, Fuc), H = 0.0;
H = HessianForOneVar(AlphaKV, UID, CID, Fuc);
if (Fuc == 0.0 && Grad <= 0.0) { Grad = 0.0; }
if (fabs(Grad) < 1e-3) { break; }
if (H == 0.0) { Fuc = 0.0; break; }
double NewtonStep = - Grad / H;
if (NewtonStep < -0.5) { NewtonStep = - 0.5; }
Fuc += NewtonStep;
if (Fuc < 0.0) { Fuc = 0.0; }
}
if (Fuc == 0.0) {
DelCom(UID, CID);
}
else {
AddCom(UID, CID, Fuc);
}
}
}
if (iter - PrevIter >= 2 * G->GetNodes() && iter > 10000) {
PrevIter = iter;
CurL = Likelihood();
if (PrevL > TFlt::Mn && ! PlotNm.Empty()) {
printf("\r%d iterations, Likelihood: %f, Diff: %f", iter, CurL, CurL - PrevL);
}
fflush(stdout);
if (CurL - PrevL <= Thres * fabs(PrevL)) { break; }
else { PrevL = CurL; }
}
}
if (! PlotNm.Empty()) {
printf("\nMLE for Lambda completed with %d iterations(%s)\n", iter, ExeTm.GetTmStr());
TGnuPlot::PlotValV(IterLV, PlotNm + ".likelihood_Q");
}
return iter;
}
