Merge branch 'master' of https://github.com/snap-stanford/snap

Credits.txt

@@ -25,6 +25,8 @@ Nikhil Khadke
 Vikesh Khanna
 Emmanouil Kiagias
 Myunghwan Kim
+Vid Kocijan
+Saachi Jain
 Jason Jong
 Peter Lofgren
 Stephen Macke
@@ -33,6 +35,7 @@ Julian McAuley
 Morteza Milani
 Riyad Parvez
 Yonathan Perez
+Mohana Prasad
 Rohan Puttagunta
 Martin Raison
 Sheila Ramaswamy
@@ -42,13 +45,15 @@ Davide Sardina
 Christoph Schaefer
 Pararth Shah
 Nicholas Shelly
+Klemen Simonic
 Guo Tai
 Ming Han Teh
 Ba-Quan Truong
 Utkarsh Upadhyay
 Theodore Vasiloudis
 Viswajith Venugopal
 Jaewon Yang
+Hao Yin
 David Zeng
 Chenguang Zhu
 Ziga Zupanec

Makefile.config

@@ -35,7 +35,7 @@ ifeq ($(UNAME), Linux)
 else ifeq ($(UNAME), Darwin)
   # OS X flags
   CC = g++
-  CXXFLAGS += -std=c++98 -Wall
+  CXXFLAGS += -std=c++98 -Wall -Wno-unknown-pragmas
   CXXFLAGS += -O3 -DNDEBUG
   CLANG := $(shell g++ -v 2>&1 | grep clang | cut -d " " -f 2)
   ifneq ($(CLANG), LLVM)

ReadMe.txt

@@ -120,6 +120,8 @@ Description of examples:
   	Estimates Kronecker graph parameter matrix.
   krongen : 
   	Generates Kronecker graphs.
+  localmotifcluster :
+	Implements a local method for motif-based clustering using MAPPR.
   lshtest :
 	Implements locality sensitive hashing.
   magfit :

Release.txt

@@ -3,6 +3,75 @@ performance system for graph and network manipulation and analysis
 
 This file contains a high-level description of changes in SNAP releases.
 
+Release 4.0, Jul 27, 2017
+
+This is a major release and brings significant new functionality: local
+higher-order clustering, counting of temporal motifs, node embedding with
+node2vec, locality sensitive hashing, K nearest neighbor graph construction,
+fast object loading using memory mapped files, and sparse attributes.
+
+Here is a detailed list:
+
+- implemented local higher-order clustering, see examples/localmotifcluster
+- implemented counting of temporal motifs, see examples/temporalmotifs
+- implemented node embedding with node2vec, see examples/node2vec
+- implemented locality sensitive hashing, see examples/lshtest
+- implemented K nearest neighbor graph construction, see
+  examples/knnjaccardsim
+- implemented methods for computing biased random walks, required for node2vec
+- implemented fast loading of objects using a memory mapped files via shared
+  memory for the following classes: TVec, TStrPool, TBigStrPool, THashKeyDat,
+  THash, TStrHash, TUNGraph, TNGraph, TMMNet, TNEANet, TNodeNet, TNodeEDatNet,
+  TNodeEdgeNet, TUndirNet, TDirNet, TTable, TTableContext
+- implemented a shared-memory class TShMIn for input stream objects
+- implemented sparse attributes (vector rather than hash based) in TMMNet and
+  TNEANet
+- implemented a backward compatible load function TNEANet::Load_V2()
+- implemented a method to convert a dense network to sparse
+  TNEANet::ConvertToSparse()
+- implemented node2vec for TNEANet, TNGraph, and TNodeEDatNet
+- defined a number of new vector and hash classes
+- implemented base TNum class for integer based classes
+- implemented a class for 64-bit integers TInt64
+- expanded TVVVec to work with 64-bit integers
+- expanded TVVec to work with 64-bit integers
+- changed API for centrality calculations so that IsDir is the last parameter
+- improved the code for betweenness centrality for directed graphs
+- implemented AddNodeUnchecked() for TNodeNet, TNodeEDatNet, TNodeEdgeNet,
+  TNEANet, TUndirNet, TDirNet
+- implemented SortNodeAdjV() for TUndirNet, TDirNet
+- implemented TAttr::Load() to work with sparse attributes
+- fixed a division by 0 in TTable::FillBucketsByWindow()
+- fixed a bug in GenRewire() for directed graphs
+- fixed a problem in GetGraphAnf(), dist=1 was redundant
+- fixed a negative index error in casc.cpp
+- fixed incorrect GCC_ATOMIC to better support ToNetwork() on Windows
+- fixed a possible infinite loop in GetRndEdgeNonAdjNode()
+- fixed node2vec crashing on directed graphs
+- fixed and optimized TMMNet::ToNetworkMP
+- added assertions for preventing writing to read only objects
+- added a test for GenRwire()
+- added tests for fast loading from shared memory
+- added tests for jaccard similarity
+- improved printing of 64-bit values
+- renamed AddAtm() to AddMP()
+- replaced GetMn() with explicit code to enable compilation in Snap.py
+- removed profiling code from GetTriangleCnt()
+- removed a redundant openmp flag from Makefile
+- removed dependency of table.h on centr.h
+- moved MapPageRank() and MapHits() from table.h to centr.h
+- moved GetMapPageRank() and GetMapHitsIterator() from table.h to centr.cpp
+- made edits for the code to compile with gcc on Mac OS X 10.9 and clang on
+  macOS 10.12
+- updated code for compilation of Snap.py on Windows
+- commented out obsolete GetMergeSortedV()
+- commented out CountTriangles(), a previous implementation of
+  GetTriangleCnt()
+- commented out dummy definitions of TNGraphMP, TNEANetMP
+- fixed a few documentation typos in alg.h
+- fixed a minor typo in agm comments
+- edited ReadMe.txt to use only ASCII characters
+
 Release 3.0, Jul 18, 2016
 
 This is a mega large release. It brings major new functionality: relational

examples/Makefile

@@ -4,12 +4,12 @@
 
 all: MakeAll
 
-MakeAll:
+MakeAll:
 	$(MAKE) -C agmfit
 	$(MAKE) -C agmgen
 	$(MAKE) -C bigclam
-	$(MAKE) -C cascades
 	$(MAKE) -C cascadegen
+	$(MAKE) -C cascades
 	$(MAKE) -C centrality
 	$(MAKE) -C cesna
 	$(MAKE) -C circles
@@ -23,13 +23,16 @@ MakeAll:
 	$(MAKE) -C graphhash
 	$(MAKE) -C infopath
 	$(MAKE) -C kcores
+	$(MAKE) -C knnjaccardsim
 	$(MAKE) -C kronem
 	$(MAKE) -C kronfit
 	$(MAKE) -C krongen
+	$(MAKE) -C localmotifcluster
+	$(MAKE) -C lshtest
 	$(MAKE) -C magfit
 	$(MAKE) -C maggen
 	$(MAKE) -C mkdatasets
-	# 'motifcluster' compilation skipped by default, it requires ARPACK
+	# 'motifcluster' compilation skipped by default, it requires ARPACK
 	#$(MAKE) -C motifcluster
 	$(MAKE) -C motifs
 	$(MAKE) -C ncpplot
@@ -41,13 +44,13 @@ MakeAll:
 	$(MAKE) -C rolx
 	$(MAKE) -C temporalmotifs
 	$(MAKE) -C testgraph
-
+
 clean:
 	$(MAKE) clean -C agmgen
 	$(MAKE) clean -C agmfit
 	$(MAKE) clean -C bigclam
-	$(MAKE) clean -C cascades
 	$(MAKE) clean -C cascadegen
+	$(MAKE) clean -C cascades
 	$(MAKE) clean -C centrality
 	$(MAKE) clean -C cesna
 	$(MAKE) clean -C circles
@@ -61,9 +64,12 @@ clean:
 	$(MAKE) clean -C graphhash
 	$(MAKE) clean -C infopath
 	$(MAKE) clean -C kcores
+	$(MAKE) clean -C knnjaccardsim
 	$(MAKE) clean -C kronem
-	$(MAKE) clean -C krongen
 	$(MAKE) clean -C kronfit
+	$(MAKE) clean -C krongen
+	$(MAKE) clean -C localmotifcluster
+	$(MAKE) clean -C lshtest
 	$(MAKE) clean -C magfit
 	$(MAKE) clean -C maggen
 	$(MAKE) clean -C mkdatasets

examples/knnjaccardsim/Makefile

@@ -0,0 +1,11 @@
+#
+#	Makefile for this SNAP example
+#	- modify Makefile.ex when creating a new SNAP example
+#
+#	implements:
+#		all (default), clean
+#
+
+include ../../Makefile.config
+include Makefile.ex
+include ../Makefile.exmain

examples/knnjaccardsim/Makefile.ex

@@ -0,0 +1,8 @@
+#
+#	configuration variables for the example
+
+## Main application file
+MAIN = knnjaccardsim
+DEPH = 
+DEPCPP = 
+

examples/knnjaccardsim/ReadMe.txt

@@ -0,0 +1,26 @@
+========================================================================
+    knnjaccardsim: KNN Jaccard Similarity
+========================================================================
+
+The example finds K nearest neighbor for every source node in the graph based
+on jaccard similarity and returns it as a PNEANet graph. 
+Input: a table, src & dst node indexes in the table schema such that they
+       form a bipartite graph, K.
+Node format in the input file name is as follows:
+  <src> <dst> 
+<src> is the source node, <dst> is the destination node.
+They form a bipartite graph.
+
+/////////////////////////////////////////////////////////////////////////////
+Parameters:
+
+/////////////////////////////////////////////////////////////////////////////
+Usage: ./knnjaccardsim <filename> <K>
+
+/////////////////////////////////////////////////////////////////////////////
+Output: three columns giving edges with the K nearest neighbors and the
+  edge similarity score
+
+Example:
+  knnjaccardsim reality.txt 3
+

examples/knnjaccardsim/knnjaccardsim.cpp

@@ -0,0 +1,39 @@
+#include "stdafx.h"
+#include "Snap.h"
+#include<iostream>
+using namespace std;
+
+const std::string currentDateTime() {
+  time_t     now = time(0);
+  struct tm  tstruct;
+  char       buf[80];
+  tstruct = *localtime(&now);
+  // Visit http://en.cppreference.com/w/cpp/chrono/c/strftime
+  // for more information about date/time format
+  strftime(buf, sizeof(buf), "%Y-%m-%d.%X", &tstruct);
+
+  return buf;
+}
+
+int main(int argc,char* argv[]) {
+  TTableContext Context;
+  //Create schema
+  //Input File Format Source,Dest,Start_Time,Duration
+  Schema TimeS;
+  TimeS.Add(TPair<TStr,TAttrType>("Source",atInt));
+  TimeS.Add(TPair<TStr,TAttrType>("Dest",atInt));
+  PTable P = TTable::LoadSS(TimeS,argv[1],&Context,' ');
+  int K = atoi(argv[2]);
+  cerr<<"Table Loaded "<<currentDateTime()<<endl;
+  PNGraph G = GetBiGraph(P, 0, 1);
+  cerr<<"Graph Generated "<<currentDateTime()<<endl;
+  #ifdef GCC_ATOMIC
+  PNEANet KNN =  KNNJaccardParallel(G, K);
+
+  for (TNEANet::TEdgeI EI = KNN->BegEI(); EI < KNN->EndEI(); EI++ ){
+    cout<<EI.GetSrcNId()<<" "<<EI.GetDstNId()<<" "<<KNN->GetFltAttrDatE(EI.GetId(), "sim")<<endl;
+    }
+  #endif
+  return 0;
+}
+

examples/knnjaccardsim/reality.txt

@@ -0,0 +1,8 @@
+1 10
+1 11
+2 10
+2 11
+2 12
+3 11
+3 12
+4 10

examples/knnjaccardsim/stdafx.cpp

@@ -0,0 +1,8 @@
+// stdafx.cpp : source file that includes just the standard includes
+// bigclam.pch will be the pre-compiled header
+// stdafx.obj will contain the pre-compiled type information
+
+#include "stdafx.h"
+
+// TODO: reference any additional headers you need in STDAFX.H
+// and not in this file

examples/knnjaccardsim/stdafx.h

@@ -0,0 +1,5 @@
+#pragma once
+
+#include "targetver.h"
+
+#include "Snap.h"

examples/knnjaccardsim/targetver.h

@@ -0,0 +1,13 @@
+#pragma once
+
+// The following macros define the minimum required platform.  The minimum required platform
+// is the earliest version of Windows, Internet Explorer etc. that has the necessary features to run 
+// your application.  The macros work by enabling all features available on platform versions up to and 
+// including the version specified.
+
+// Modify the following defines if you have to target a platform prior to the ones specified below.
+// Refer to MSDN for the latest info on corresponding values for different platforms.
+#ifndef _WIN32_WINNT            // Specifies that the minimum required platform is Windows Vista.
+#define _WIN32_WINNT 0x0600     // Change this to the appropriate value to target other versions of Windows.
+#endif
+