void MakeSignEpinions() {
TSsParser Ss("/u/ana/data/EpinionRatings/user_rating.txt", ssfTabSep);
//PSignNet Net = TSignNet::New();
TPt<TNodeEDatNet<TInt, TInt> > Net = TNodeEDatNet<TInt, TInt>::New();
TStrHash<TInt> StrSet(Mega(1), true);
while (Ss.Next()) {
if ( ((TStr)Ss[0]).IsPrefix("#") )
continue;
const int SrcNId = StrSet.AddKey(Ss[0]);
const int DstNId = StrSet.AddKey(Ss[1]);
if (! Net->IsNode(SrcNId)) { Net->AddNode(SrcNId); }
if (! Net->IsNode(DstNId)) { Net->AddNode(DstNId); }
const int Sign = ((TStr)Ss[2]).GetInt();
Net->AddEdge(SrcNId, DstNId, Sign);
}
// PrintGraphStatTable(Graph, OutFNm, Desc);
TStr OutFNm = "soc-sign-epinions-user-ratings";
TStr Desc = "Epinions signed social network";
// copied from gio.h - line 111
FILE *F = fopen(OutFNm.CStr(), "wt");
fprintf(F, "# Directed graph: %s\n", OutFNm.CStr());
if (! Desc.Empty())
fprintf(F, "# %s\n", (Desc).CStr());
fprintf(F, "# Nodes: %d Edges: %d\n", Net->GetNodes(), Net->GetEdges());
fprintf(F, "# FromNodeId\tToNodeId\tSign\n");
for (TNodeEDatNet<TInt,TInt>::TEdgeI ei = Net->BegEI(); ei < Net->EndEI(); ei++) {
fprintf(F, "%d\t%d\t%d\n", ei.GetSrcNId(), ei.GetDstNId(), ei()());
}
fclose(F);
PrintGraphStatTable(Net, OutFNm, Desc);
}
// Slashdot network
void MakeSlashdotNet(TStr InFNm, TStr OutFNm, TStr Desc) {
TSsParser Ss(InFNm, ssfTabSep);
PNGraph Graph = TNGraph::New();
TStrHash<TInt> StrSet(Mega(1), true);
while (Ss.Next()) {
const int SrcNId = StrSet.AddKey(Ss[0]);
if (! Graph->IsNode(SrcNId)) { Graph->AddNode(SrcNId); }
for (int dst = 2; dst < Ss.Len(); dst++) {
const int DstNId = StrSet.AddKey(Ss[dst]);
if (! Graph->IsNode(DstNId)) { Graph->AddNode(DstNId); }
Graph->AddEdge(SrcNId, DstNId);
}
}
PrintGraphStatTable(Graph, OutFNm, Desc);
}
void MakeLJNets(TStr InFNm, TStr OutFNm, TStr Desc){
TStrHash<TInt> StrSet(Mega(1), true);
for (int i = 1; i < 13; i++){
TStr tmp = "";
if (i < 10)
tmp = InFNm + "ljgraph.0" + TInt::GetStr(i);
else
tmp = InFNm + "ljgraph." + TInt::GetStr(i);
printf("%s\n",tmp());
TSsParser Ss(tmp, ssfTabSep);
PNGraph Graph = TNGraph::New();
while (Ss.Next()) {
const int SrcNId = StrSet.AddKey(Ss[0]);
if (! Graph->IsNode(SrcNId)) { Graph->AddNode(SrcNId); }
for (int dst = 2; dst < Ss.Len(); dst++) {
TStr ls,rs;
((TStr)Ss[dst]).SplitOnCh(ls,' ',rs);
if (ls == ">"){
const int DstNId = StrSet.AddKey(rs);
if (! Graph->IsNode(DstNId)) { Graph->AddNode(DstNId); }
Graph->AddEdge(SrcNId, DstNId);
}
}
}
if (i < 10)
OutFNm = "soc-lj-friends.0"+TInt::GetStr(i);
else
OutFNm = "soc-lj-friends."+TInt::GetStr(i);
PrintGraphStatTable(Graph, OutFNm, Desc);
}
}
void MakeLJGroupsNets(const TStr InFNm, TStr OutFNm, TStr Desc){
TStrHash<TInt> StrSetU(Mega(1), true);
TStrHash<TInt> StrSetG(Mega(1), true);
for (int i = 1; i < 13; i++){
TStr tmp = "";
if (i < 10)
tmp = InFNm + "ljgraph.0" + TInt::GetStr(i);
else
tmp = InFNm + "ljgraph." + TInt::GetStr(i);
printf("%s\n",tmp());
TSsParser Ss(tmp, ssfTabSep);
PNGraph Graph = TNGraph::New();
while (Ss.Next()) {
const int SrcNId = StrSetU.AddKey(Ss[0]);
if (! Graph->IsNode(SrcNId))
Graph->AddNode(SrcNId);
for (int dst = 2; dst < Ss.Len(); dst++) {
TStr ls,rs;
((TStr)Ss[dst]).SplitOnCh(ls,' ',rs);
if (ls == ">"){
const int DstNId = StrSetU.AddKey(rs);
if (! Graph->IsNode(DstNId)) { Graph->AddNode(DstNId); }
Graph->AddEdge(SrcNId, DstNId);
}
}
}
if (i < 10)
OutFNm = "soc-lj-friends.0"+TInt::GetStr(i);
else
OutFNm = "soc-lj-friends."+TInt::GetStr(i);
// PrintGraphStatTable(Graph, OutFNm, Desc);
}
for (int i = 5; i < 14; i++){
TStr tmp = "";
tmp = InFNm + "ljcomm." + TInt::GetStr(i);
printf("%s\n",tmp());
TSsParser Ss1(tmp, ssfTabSep);
PNGraph Graph1 = TNGraph::New();
PNGraph Graph2 = TNGraph::New();
TSsParser Ss(tmp, ssfTabSep);
while (Ss.Next()) {
const int SrcNId = StrSetU.AddKey(Ss[0]);
if (! Graph1->IsNode(SrcNId)) { Graph1->AddNode(SrcNId); }
if (! Graph2->IsNode(SrcNId)) { Graph2->AddNode(SrcNId); }
for (int dst = 2; dst < Ss.Len(); dst++) {
TStr ls,rs;
((TStr)Ss[dst]).SplitOnCh(ls,' ',rs);
const int DstNId = StrSetG.AddKey(rs) + 10000000;
if (! Graph1->IsNode(DstNId)) { Graph1->AddNode(DstNId); }
if (! Graph2->IsNode(DstNId)) { Graph2->AddNode(DstNId); }
if (ls == "<") // member
Graph1->AddEdge(SrcNId, DstNId);
else if (ls == ">") // watching
Graph2->AddEdge(SrcNId, DstNId);
}
}
OutFNm = "soc-lj-comm-";
TStr s = "";
if (i < 10)
s = "0";
PrintGraphStatTable(Graph1, OutFNm+"members-" + s + TInt::GetStr(i), Desc+" communities - members");
PrintGraphStatTable(Graph2, OutFNm+"watchers-"+ s + TInt::GetStr(i), Desc+" communities - watchers");
}
}
// Parse files downloaded from IMDB. Actors point to movies.
// Files: actors.list.gz, languages.list.gz, countries.list.gz
PImdbNet TImdbNet::LoadFromImdb(const TStr& DataDir) {
PImdbNet Net = TImdbNet::New();
Net->Reserve((int) Mega(2.5), -1);
// ACTORS
{ TSsParser Ss(DataDir+"\\actors.list.gz", ssfTabSep);
while (Ss.Next() && strcmp(Ss[0],"THE ACTORS LIST")!=0) { }
Ss.Next(); Ss.Next(); Ss.Next();
int ActorId = -1, NAct=0;
for (int i = 0; Ss.Next(); i++) {
//printf("%s\n", Ss.DumpStr());
int mPos = 0;
if (Ss.Len() > 1) { // actor
ActorId = Net->AddStr(Ss[0]);
if (Net->IsNode(ActorId)) {
printf(" actor '%s'(%d) is already a node %s\n", Ss[0],
ActorId, TImdbNet::GetMovieTyStr((TMovieTy) Net->GetNDat(ActorId).Type.Val).CStr());
continue;
} else { Net->AddNode(ActorId); }
TImdbNode& Node = Net->GetNDat(ActorId);
Node.Name = ActorId; NAct++;
Node.Type = mtyActor;
mPos = 1; while (strlen(Ss[mPos])==0) { mPos++; }
}
// movie (delete everything last [)
// also parse the position <>, but is a property of an edge (actor, movie) pair
char *DelFrom;
char *C1 = strrchr(Ss[mPos], '<');
char *C2 = strrchr(Ss[mPos], '[');
if (C1==NULL) { DelFrom=C2; }
else if (C2==NULL) { DelFrom=C1; }
else { DelFrom = TMath::Mn(C1, C2); }
if (DelFrom != NULL) {
char *mov = DelFrom; while (*mov) { *mov=0; mov++; }
mov = (char *) DelFrom-1; while (TCh::IsWs(*mov)) { *mov=0; mov--; }
}
const int MovNId = Net->AddStr(Ss[mPos]);
if (! Net->IsNode(MovNId)) {
Net->AddNode(MovNId);
TImdbNode& Node = Net->GetNDat(MovNId);
Node.Type = mtyMovie; Node.Year = GetYearFromTitle(Ss[mPos]);
}
if (Net->IsEdge(ActorId, MovNId)) {
printf(" already an edge %d %d\n", ActorId, MovNId); }
else { Net->AddEdge(ActorId, MovNId); }
if ((i+1) % Kilo(10) == 0) {
printf("\r %d", (i+1)/1000);
if ((i+1) % Kilo(100) == 0) {
printf(" nodes: %d, edges: %d, actors: %d",
Net->GetNodes(), Net->GetEdges(), NAct); }
}
}
printf("\n=== nodes: %d, edges: %d, actors: %d", Net->GetNodes(), Net->GetEdges(), NAct); }
// MOVIE LANGUAGE */
{ TSsParser Ss(DataDir+"\\language.list.gz", ssfTabSep);
while (Ss.Next() && strcmp(Ss[0],"LANGUAGE LIST")!=0) { }
Ss.Next();
int LangCnt=0, i;
for (i = 0; Ss.Next(); i++) {
char *Mov = Ss[0];
char *Lang = Ss[Ss.Len()-1];
if (Net->IsStr(Mov)) {
const int NId = Net->GetStrId(Mov);
Net->GetNDat(NId).Lang = Net->AddStr(Lang);
LangCnt++;
} //else { printf("movie not found: '%s'\n", Mov); }
if ((i+1) % Kilo(10) == 0) {
printf("\r %d found %d ", (i+1), LangCnt); }
}
printf("\n LANG: total movies: %d, found %d\n", (i+1), LangCnt); }
// MOVIE COUNTRY
{ TSsParser Ss(DataDir+"\\countries.list.gz", ssfTabSep);
while (Ss.Next() && strcmp(Ss[0],"COUNTRIES LIST")!=0) { }
Ss.Next();
int LangCnt=0, i;
for (i = 0; Ss.Next(); i++) {
char *Mov = Ss[0];
char *Cntry = Ss[Ss.Len()-1];
if (Net->IsStr(Mov)) {
const int NId = Net->GetStrId(Mov);
Net->GetNDat(NId).Cntry = Net->AddStr(Cntry);
LangCnt++;
} //else { printf("country not found: '%s'\n", Mov); }
if ((i+1) % Kilo(10) == 0) {
printf("\n %d found %d ", (i+1), LangCnt); }
}
printf("\r CNTRY: total movies: %d, found %d\n", (i+1), LangCnt); }
return Net;
}
void BigMain(int argc, char* argv[]) {
TExeTm ExeTm;
Env = TEnv(argc, argv, TNotify::StdNotify);
Env.PrepArgs("QuotesApp");
const TStr ToDo = Env.GetIfArgPrefixStr("-do:", "", "To do").GetLc();
if (Env.IsEndOfRun()) {
printf("To do:\n");
printf(" MkDataset : Make memes dataset (extract quotes and save txt)\n");
printf(" ExtractSubset : Extract a subset of memes containing particular words\n");
printf(" MemesToQtBs : Load memes dataset and create quote base\n");
printf(" MkClustNet : Build cluster network from the quote base\n");
return;
}
#pragma region mkdataset
// extract quotes and links and make them into a single file
if (ToDo == "mkdataset") {
const TStr InFNm = Env.GetIfArgPrefixStr("-i:", "files.txt", "Spinn3r input files (one file per line)");
const TStr OutFNm = Env.GetIfArgPrefixStr("-o:", "Spinn3r-dataset.txt", "Output file");
const int MinQtWrdLen = Env.GetIfArgPrefixInt("-w:", 3, "Minimum quote word length");
const TStr UrlFNm = Env.GetIfArgPrefixStr("-u:", "", "Seen url set (THashSet<TMd5Sig>) file name");
const bool UrlOnlyOnce = Env.GetIfArgPrefixBool("-q:", true, "Only keep unique Urls");
//// parse directly from Spinn3r
TStr Spinn3rFNm;
THashSet<TMd5Sig> SeenUrlSet;
if (UrlOnlyOnce && ! UrlFNm.Empty()) { // keep track of already seen urls (so that there are no duplicate urls)
TFIn FIn(UrlFNm); SeenUrlSet.Load(FIn);
}
FILE *F = fopen(OutFNm.CStr(), "wt");
TFIn FIn(InFNm);
int Items=0;
for (int f=0; FIn.GetNextLn(Spinn3rFNm); f++) {
TQuoteExtractor QE(Spinn3rFNm.ToTrunc());
printf("Processing %02d: %s [%s]\n", f+1, Spinn3rFNm.CStr(), TExeTm::GetCurTm());
fflush(stdout);
for (int item = 0; QE.Next(); item++) {
const TMd5Sig PostMd5(QE.PostUrlStr);
if (QE.QuoteV.Empty() && QE.LinkV.Empty()) { continue; } // no quotes, no links
if (UrlOnlyOnce) {
if (SeenUrlSet.IsKey(PostMd5)) { continue; }
SeenUrlSet.AddKey(PostMd5);
}
fprintf(F, "P\t%s\n", QE.PostUrlStr.CStr());
//if (QE.PubTm > TSecTm(2008,8,30) || QE.PubTm < TSecTm(2008,7,25)) { printf("%s\n", QE.PubTm.GetStr().CStr()); }
fprintf(F, "T\t%s\n", QE.PubTm.GetYmdTmStr().CStr());
for (int q = 0; q < QE.QuoteV.Len(); q++) {
if (TStrUtil::CountWords(QE.QuoteV[q]) >= MinQtWrdLen) {
fprintf(F, "Q\t%s\n", QE.QuoteV[q].CStr()); }
}
for (int l = 0; l < QE.LinkV.Len(); l++) {
fprintf(F, "L\t%s\n", QE.LinkV[l].CStr()); }
fprintf(F, "\n");
if (item>0 && item % Kilo(100) == 0) {
QE.DumpStat(); QE.ExeTm.Tick(); }
Items++;
}
printf("file done. Total %d all posts, %d all items\n", SeenUrlSet.Len(), Items);
fflush(stdout);
}
printf("all done. Saving %d post urls\n", SeenUrlSet.Len()); fflush(stdout);
if (! SeenUrlSet.Empty()) {
TFOut FOut(OutFNm.GetFMid()+".SeenUrlSet");
SeenUrlSet.Save(FOut);
}
fclose(F);
}
#pragma endregion mkdataset
#pragma region extractsubset
// save posts with memes containing particular words
else if (ToDo == "extractsubset") {
const TStr InFNmWc = Env.GetIfArgPrefixStr("-i:", "memes_*.rar", "Input file prefix");
const bool IsInFNmWc = Env.GetIfArgPrefixBool("-w:", true, "Input is wildcard (else a file with list of input files)");
const TStr OutFNm = Env.GetIfArgPrefixStr("-o:", "memes-subset.txt", "Output memes file");
const TStr WordsFNm = Env.GetIfArgPrefixStr("-p:", "phrases-in.txt", "Phrases that memes have to contain");
TChAV CatchMemeV;// = TStr::GetV("great depression", "economic meltdown", "recession had bottomed out", "green shoots", "slow recovery", "gradual recovery");
printf("Loading %s\n", WordsFNm.CStr());
{ TFIn FIn(WordsFNm);
for (TStr Ln; FIn.GetNextLn(Ln); ) {
printf(" %s\n", Ln.GetLc().CStr());
CatchMemeV.Add(Ln.GetLc()); }
}
printf("%d strings loaded\n", CatchMemeV.Len());
TFOut FOut(OutFNm);
TMemesDataLoader Memes(InFNmWc, IsInFNmWc);
for (int posts = 0, nsave=0; Memes.LoadNext(); posts++) {
bool DoSave = false;
for (int m = 0; m < Memes.MemeV.Len(); m++) {
for (int i = 0; i < CatchMemeV.Len(); i++) {
if (Memes.MemeV[m].SearchStr(CatchMemeV[i]) != -1) {
DoSave=true; break; }
}
if (DoSave) { break; }
}
if (DoSave) { Memes.SaveTxt(FOut); nsave++; }
if (posts % Mega(1) == 0) {
printf("%dm posts, %d saved\n", posts/Mega(1), nsave);
FOut.Flush();
}
}
}
#pragma endregion extractsubset
#pragma region memestoqtbs
// load memes dataset (MkDataset) and create quote base
else if (ToDo == "memestoqtbs") {
const TStr InFNm = Env.GetIfArgPrefixStr("-i:", "201007_201107.txt", "Input Memes dataset files");
const TStr MediaUrlFNm = Env.GetIfArgPrefixStr("-u:", "news_media.txt", "Fule with news media urls");
const TStr Pref = Env.GetIfArgPrefixStr("-o:", "qt", "Output file name prefix");
const int MinWrdLen = Env.GetIfArgPrefixInt("-l:", 4, "Min quote word length");
const int MinMemeFq = Env.GetIfArgPrefixInt("-f:", 5, "Min meme frequency");
const TStr MinTmStr = Env.GetIfArgPrefixStr("-mint:", "20100714", "Min time of quotes, format = YYYYMMDD");
const TStr MaxTmStr = Env.GetIfArgPrefixStr("-maxt:", "20110728", "Max time of quotes, format = YYYYMMDD");
TSecTm MinTm(atoi(MinTmStr.GetSubStr(0,3).CStr()),atoi(MinTmStr.GetSubStr(4,5).CStr()),atoi(MinTmStr.GetSubStr(6,7).CStr()));
TSecTm MaxTm(atoi(MaxTmStr.GetSubStr(0,3).CStr()),atoi(MaxTmStr.GetSubStr(4,5).CStr()),atoi(MaxTmStr.GetSubStr(6,7).CStr()));
PQuoteBs QtBs = TQuoteBs::New();
int HashTableSize = 100; // 100 for each quarter, for one year data, use 400
int UrlSetSize = 4 * HashTableSize;
QtBs->ConstructQtBsZarya(InFNm, Pref, MediaUrlFNm, MinTm, MaxTm, MinWrdLen, MinMemeFq, HashTableSize, UrlSetSize);
}
#pragma endregion memestoqtbs
#pragma region mkclustnet
// make cluster network
else if (ToDo == "mkclustnet") {
TStr InQtBsNm = Env.GetIfArgPrefixStr("-i:", "", "Input quote base file name");
TStr Pref = Env.GetIfArgPrefixStr("-o:", "qt", "Output network/updated QtBs filename");
TStr BlackListFNm = Env.GetIfArgPrefixStr("-b:", "quote_blacklist.txt", "Blacklist file name");
bool IsShglReady = Env.GetIfArgPrefixBool("-shglready:", false, "Indicate whether shingle hash table is ready");
bool IsNetReady = Env.GetIfArgPrefixBool("-netready:", false, "Indicate whether cluster net is ready");
double BktThresh = Env.GetIfArgPrefixFlt("-bktthresh:", 0.4, "Threshold for bad shingle bucket elimination");
double MxTmDelay = Env.GetIfArgPrefixFlt("-delaythresh:", 5, "Max time delay between two quotes in the same cluster");
double MxTmDev = Env.GetIfArgPrefixFlt("-devthresh:", 3, "Max time deviation for a quote to be specific rather than general");
double RefineThresh = Env.GetIfArgPrefixFlt("-refinethresh:", 0.2, "Threshold for merging quote cluster in refining process");
const int MinWrdLen = Env.GetIfArgPrefixInt("-minl:", 4, "Min quote word length");
const int MaxWrdLen = Env.GetIfArgPrefixInt("-maxl:", 200, "Max quote word length");
const int MinMemeFq = Env.GetIfArgPrefixInt("-mf:", 5, "Min meme frequency");
const int MinClustFq = Env.GetIfArgPrefixInt("-cf:", 50, "Min quote cluster frequency");
// Load quote base
PQuoteBs QtBs;
if (TZipIn::IsZipFNm(InQtBsNm)) { TZipIn ZipIn(InQtBsNm); QtBs = TQuoteBs::Load(ZipIn); }
else { TFIn FIn(InQtBsNm); QtBs = TQuoteBs::Load(FIn); }
// Cluster the quotes
QtBs->ClusterQuotes(MinMemeFq, MinWrdLen, MaxWrdLen, BlackListFNm, Pref, IsShglReady, IsNetReady, BktThresh, MxTmDelay, MxTmDev, RefineThresh);
// Dump the clusters
bool SkipUrl = true, FlashDisp = true;
QtBs->DumpQuoteClusters(MinWrdLen, MinClustFq, SkipUrl, FlashDisp, Pref);
}
#pragma endregion mkclustnet
#pragma region memeclust
else if (ToDo.SearchStr(TStr("memeclust")) >= 0) {
const TStr InFNm = Env.GetIfArgPrefixStr("-i:", "201101.txt", "Input Memes dataset files");
const TStr MediaUrlFNm = Env.GetIfArgPrefixStr("-u:", "news_media.txt", "Fule with news media urls");
TStr Pref = Env.GetIfArgPrefixStr("-o:", "qt", "Output file name prefix");
const int MinWrdLen = Env.GetIfArgPrefixInt("-minl:", 4, "Min quote word length");
const int MaxWrdLen = Env.GetIfArgPrefixInt("-maxl:", 200, "Max quote word length");
const int MinMemeFq = Env.GetIfArgPrefixInt("-f:", 5, "Min meme frequency");
const int MinClustFq = Env.GetIfArgPrefixInt("-cf:", 50, "Min quote cluster frequency");
TStr BlackListFNm = Env.GetIfArgPrefixStr("-b:", "quote_blacklist.txt", "Blacklist file name");
bool IsQtBsReady = Env.GetIfArgPrefixBool("-qtbsready:", false, "Indicate whether quote base is ready and can be loaded readily");
bool IsShglReady = Env.GetIfArgPrefixBool("-shglready:", false, "Indicate whether shingle hash table is ready");
bool IsNetReady = Env.GetIfArgPrefixBool("-netready:", false, "Indicate whether cluster net is ready");
double BktThresh = Env.GetIfArgPrefixFlt("-bktthresh:", 0.4, "Threshold for bad shingle bucket elimination");
double MxTmDelay = Env.GetIfArgPrefixFlt("-delaythresh:", 5, "Max time delay between two quotes in the same cluster");
double MxTmDev = Env.GetIfArgPrefixFlt("-devthresh:", 3, "Max time deviation for a quote to be specific rather than general");
double RefineThresh = Env.GetIfArgPrefixFlt("-refinethresh:", 0.2, "Threshold for merging quote cluster in refining process");
const TStr MinTmStr = Env.GetIfArgPrefixStr("-mint:", "20010101", "Min time of quotes, format = YYYYMMDD");
const TStr MaxTmStr = Env.GetIfArgPrefixStr("-maxt:", "20200101", "Max time of quotes, format = YYYYMMDD");
TSecTm MinTm(atoi(MinTmStr.GetSubStr(0,3).CStr()),atoi(MinTmStr.GetSubStr(4,5).CStr()),atoi(MinTmStr.GetSubStr(6,7).CStr()));
TSecTm MaxTm(atoi(MaxTmStr.GetSubStr(0,3).CStr()),atoi(MaxTmStr.GetSubStr(4,5).CStr()),atoi(MaxTmStr.GetSubStr(6,7).CStr()));
// Construct the quote base from Zarya data
PQuoteBs QtBs = TQuoteBs::New();
if (!IsQtBsReady) {
int HashTableSize = 100; // 100 for each quarter, for one year data, use 400
if (ToDo == "memeclustzarya") {
int UrlSetSize = 4 * HashTableSize;
QtBs->ConstructQtBsZarya(InFNm, Pref, MediaUrlFNm, MinTm, MaxTm, MinWrdLen, MinMemeFq, HashTableSize, UrlSetSize);
} else if (ToDo == "memeclustqtonly") {
QtBs->ConstructQtBsQtOnly(InFNm, Pref, MediaUrlFNm, MinWrdLen, MinMemeFq, HashTableSize);
} else if (ToDo == "memeclustqttime") {
QtBs->ConstructQtBsQtTime(InFNm, Pref, MediaUrlFNm, MinWrdLen, MinMemeFq, HashTableSize);
} else {
printf("Please specify one of the three options for -do : memeclustzarya, memeclustqtonly, memeclustqttime!\n");
return;
}
} else {
TStr InQtBsNm = TStr::Fmt("%s-w%dmfq%d.QtBs", Pref.CStr(), MinWrdLen, MinMemeFq);
if (TZipIn::IsZipFNm(InQtBsNm)) { TZipIn ZipIn(InQtBsNm); QtBs = TQuoteBs::Load(ZipIn); }
else { TFIn FIn(InQtBsNm); QtBs = TQuoteBs::Load(FIn); }
}
// Cluster the quotes
QtBs->ClusterQuotes(MinMemeFq, MinWrdLen, MaxWrdLen, BlackListFNm, Pref, IsShglReady, IsNetReady, BktThresh, MxTmDelay, MxTmDev, RefineThresh);
// Dump the clusters
bool SkipUrl = true, FlashDisp = true;
QtBs->DumpQuoteClusters(MinWrdLen, MinClustFq, SkipUrl, FlashDisp, Pref);
}
#pragma endregion memeclust
}
