post merge branch tree_stripping

README.md

@@ -20,16 +20,16 @@ These are sorted with the smallest vertex in first position, and such as the sec
 <a name="s_stat"></a>
 
 - Home page: https://github.com/skramm/udgcd
-- beta, may produce a result, no guarantee at all (this is a stalled project)
-- Author: Sebastien Kramm
+- Author: Sebastien Kramm, [email protected]
 - Latest news:
+  - 2023-07-17: (**major update**) Algorithm improvment, now much more efficient, see [details here](misc/tree_stripping.md)
   - 2023-07-03: was reported to [build on Windows using VisualStudio](https://github.com/skramm/udgcd/issues/4#issuecomment-1611426339)
   - 2020-06-09: experimental code and preliminar release, source is pretty messy, but it works fine, give it a try (instructions below).
 - Released under the Boost licence.
 
 ### Features
 
-- Single-file, header only, OS agnostic
+- Single-file header only library, OS agnostic
 - Works for graphs holding unconnected sub-graphs.
 - Works for non-planar graphs
 - Modern C++ design (RAII), basic C++11.
@@ -75,7 +75,7 @@ This will copy the file in `/usr/local/include/`
 Some additional apps are included, that are build by the makefile:
  - `read_graph.cpp`: reads graph from a text file given as argument, computes its cycles, prints them and generate the corresponding dot file.
  - `random_test.cpp`: generates a random graph computes its cycles, prints them and generate the corresponding dot file.
- - `sample_?.cpp`: c++ apps that build a graph and computes its cycles.
+ - `sample_?.cpp`: C++ apps that build a graph and computes its cycles.
 
 #### Build options:
  - The provided makefile is not requested to use the library, as it is "header-only".
@@ -145,10 +145,16 @@ This step is the most time-consuming.
 - The third steps does some post-processing:
 sort cycles by decreasing length, and do Gaussian Elimination to retain a Minimal Cycle Basis (MCB).
 
+### Dependencies
+
+- boost::graph - https://www.boost.org/doc/libs/1_82_0/libs/graph/doc/
+(warning: boost graph itself has a lot of other boost dependencies)
+- boost::dynamic_bitset https://www.boost.org/doc/libs/1_82_0/libs/dynamic_bitset/dynamic_bitset.html
+
+
 ### References
 <a name="s_ref"></a>
 
-- BGL: http://www.boost.org/doc/libs/1_59_0/libs/graph/doc
 - https://en.wikipedia.org/wiki/Cycle_basis
 - J. D. Horton, <i>A polynomial-time algorithm to find a shortest cycle basis of a graph</i>, SIAM Journal of Computing 16, 1987, pp. 359–366
 [link](https://epubs.siam.org/doi/10.1137/0216026).

demo/common_sample.h

@@ -32,9 +32,16 @@ Home page: https://github.com/skramm/udgcd
 #include <string>
 #include <numeric>
 
+#ifdef UDGCD_USE_M4RI
+	#define HAS_M4RI "YES\n"
+#else
+	#define HAS_M4RI "NO\n"
+#endif
+
 #define SHOW_INFO \
 	std::cout << "-START: " << __FILE__ \
-		<< "\n-built with Boost " << BOOST_LIB_VERSION << '\n'
+		<< "\n-built with Boost " << BOOST_LIB_VERSION \
+		<< "\n-build option: USE_M4RI: " << HAS_M4RI << "\n";
 
 extern std::string prog_id; // allocated in samples files
 int g_idx = 0;
@@ -473,7 +480,7 @@ The counterpart is that we do not read the vertices/edges properties that can be
 
 \warning This is a minimal reader, don't expect any fancy features.
 
-Limitation/features:  as we do not handle labels, the index in input file will be the BGL vertex index.<br>
+Limitation/features: as we do not handle labels, the index in input file will be the BGL vertex index.<br>
 Drawback: if a node is missing in input file, it will be in the generated graph, because BGL indexes are always
 consecutives.
 
@@ -497,7 +504,7 @@ will generate a graph of 6 vertices (0 to 5), with only 0 and 1 connected.
 */
 template<typename graph_t>
 graph_t
-loadGraph_dot( const char* fname )
+loadGraph_dot( std::string fname )
 {
 	std::cout<< " - Reading file:" << fname << '\n';
 	std::ifstream f( fname );
@@ -641,7 +648,7 @@ loadGraph_dot( const char* fname )
 /// Load graph from custom simple text format
 template<typename graph_t>
 graph_t
-loadGraph_txt( const char* fname )
+loadGraph_txt( std::string fname )
 {
 	graph_t g;
 
@@ -734,42 +741,51 @@ The first value is 0 in case of success, -1 if incorrect cycles were found.
 If other value, it is the absolute number of differences between:
 - the \b computed number of cycles
 - and the \b expected number of cycles
-
 */
 template<typename graph_t,typename vertex_t>
 std::pair<int,std::vector<std::vector<vertex_t>>>
-processGraph( graph_t& g )
+processGraph( graph_t& g, const udgcd::RunTimeOptions& rtOptions )
 {
 	auto expected = printGraphInfo( g );
 
 	udgcd::UdgcdInfo info;
+	info.runTime = rtOptions;
 	auto cycles = udgcd::findCycles<graph_t,vertex_t>( g, info );
-//	udgcd::printPaths( std::cout, cycles, "final" );
+
 	if( expected != cycles.size() )
 		std::cout << "ERROR: computed nb of cycles is not what expected (expected=" << expected << ")\n";
 
 //	std::cout << "diff=" << (int)expected - (int)cycles.size() << "\n";
 
 	udgcd::priv::printStatus( std::cout, cycles, __LINE__ );
 
-	auto check = udgcd::priv::checkCycles( cycles, g );
-	if( check.first != 0 )
-	{
-		std::cout << "ERROR: " << check.first << " incorrect cycles found\n";
-		return std::make_pair(-1, cycles );
-	}
-	if( check.second != 0 )
+	if( rtOptions.printCycles )
+		udgcd::printPaths( std::cout, cycles, "final" );
+
+	if( rtOptions.doChecking )
 	{
-		std::cout << "Found: " << check.second << " non chordless cycles\n";
+		auto check = udgcd::priv::checkCycles( cycles, g );
+		if( check.first != 0 )
+		{
+			std::cout << "ERROR: " << check.first << " incorrect cycles found\n";
+			return std::make_pair(-1, cycles );
+		}
+		if( check.second != 0 )
+		{
+			std::cout << "Found: " << check.second << " non chordless cycles\n";
+		}
 	}
 
 	info.print( std::cout );
 //	info.printCSV( std::cerr );
 
-	std::cout << "Histogram of cycle sizes:\n";
-	auto histog = buildSizeHistogram( cycles );
-	for( size_t i=0; i<histog.size(); i++ )
-		std::cout << i+3 << ":" << histog[i] << "\n";
+	if( rtOptions.printHistogram )
+	{
+		std::cout << "Histogram of cycle sizes:\n";
+		auto histog = buildSizeHistogram( cycles );
+		for( size_t i=0; i<histog.size(); i++ )
+			std::cout << i+3 << ":" << histog[i] << "\n";
+	}
 
 	auto diff = (int)cycles.size() - (int)expected;
 	return std::make_pair(diff, cycles );

demo/random_test.cpp

@@ -113,7 +113,8 @@ int main( int argc, const char** argv )
 	SaveGraph( g, std::to_string(current_time) );
 	sample::renderGraph( g, "gen_" + std::to_string(current_time) );
 
-	auto result = sample::processGraph<graph_t,vertex_t>( g );
+	udgcd::RunTimeOptions rtOptions;
+	auto result = sample::processGraph<graph_t,vertex_t>( g, rtOptions );
 
 	return result.first;
 }

demo/read_graph.cpp

@@ -7,7 +7,9 @@
 \file read_graph.cpp
 \brief Read a graph in a file, searches for cycles, and make sure these are correct.
 
-Also generates a .dot file that can be graphically rendered with Graphviz
+Also generates a .dot file in the `out/` folder, that can be graphically rendered with Graphviz (done with `$ make svg`).
+
+Has some runtime switches, see help()
 */
 
 #include <iostream>
@@ -18,6 +20,17 @@ Also generates a .dot file that can be graphically rendered with Graphviz
 std::string prog_id = "read_graph";
 #include "common_sample.h"
 
+/// help function
+/// \sa RunTimeOptions
+void help()
+{
+	std::cout << "Options:" \
+		"\n -t: print cycles as trees" \
+		"\n -p: print produced cycles" \
+		"\n -h: print histogram of cycles length" \
+		"\n -c: does a checking of correctness of computed cycles" \
+		"\n";
+}
 
 //-------------------------------------------------------------------
 /// see read_graph.cpp
@@ -34,55 +47,71 @@ int main( int argc, const char** argv )
 
 	using vertex_t = boost::graph_traits<graph_t>::vertex_descriptor;
 
-#ifdef UDGCD_USE_M4RI
-	std::cout << "Build using libM4ri\n";
-#endif
-
 	if( argc < 2 )
 	{
-		std::cout << "missing input filename, exit.\n";
+		std::cout << "missing input filename, exit\n";
+		help();
 		return 1;
 	}
-	auto vs = sample::splitString( argv[1], '.' );
+	std::string fname = argv[argc-1];
+	auto vs = sample::splitString( fname, '.' );
 	if( vs.size() < 2 )
 	{
-		std::cerr << "Error, input file '" <<  argv[1] << "' has no extension\n";
+		std::cerr << "Error, input file '" << fname << "' has no extension\n";
 		return 1;
 	}
 
 	graph_t gr;
 	if( vs.back() == "dot" )
 	{
-		gr = sample::loadGraph_dot<graph_t>( argv[1] );
+		gr = sample::loadGraph_dot<graph_t>( fname );
 	}
 	else
 	{
 		if( vs.back() == "txt" )
 		{
-			gr = sample::loadGraph_txt<graph_t>( argv[1] );
+			gr = sample::loadGraph_txt<graph_t>( fname );
 		}
 		else
 		{
-			std::cerr << "Error, input file '" <<  argv[1] << "' extension invalid\n";
+			std::cerr << "Error, input file '" <<  fname << "' extension invalid\n";
 			return 1;
 		}
 	}
 
 
-	auto vs1 = sample::splitString( argv[1], '/' );      // splits by '/', and keep the last one (filename)
+	auto vs1 = sample::splitString( fname, '/' );      // splits by '/', and keep the last one (filename)
 	auto vs2 = sample::splitString( vs1.back(), '.' );     // splits by the point (if any)
 	sample::renderGraph( gr, vs2[0] );
 
+	udgcd::RunTimeOptions rtOptions;
+
 	bool noProcess(false);
+	bool verbose(false);
 	if( argc > 2 )
 	{
-		std::string a( argv[2] );
-		if( a == "-n" )
-			noProcess = true;
+		for( int i=1;i<argc-1;i++ )
+		{
+			std::string a( argv[i] );
+			if( a == "-n" )
+				noProcess = true;
+			if( a == "-v" )
+				verbose = true;
+			if( a == "-t" )
+				rtOptions.printTrees = true;
+			if( a == "-p" )
+				rtOptions.printCycles = true;
+			if( a == "-h" )
+				rtOptions.printHistogram = true;
+			if( a == "-c" )
+				rtOptions.doChecking = true;
+
+		}
 	}
 	if( noProcess )
 		return 0;
-	auto result = sample::processGraph<graph_t,vertex_t>( gr );
+
+	auto result = sample::processGraph<graph_t,vertex_t>( gr, rtOptions );
 	sample::renderGraph2( gr, result.second, vs2[0]+"_color" );
 	return result.first;
 }

makefile

@@ -56,6 +56,7 @@ APP=udgcd.hpp
 HEADER=udgcd.hpp
 HEADERS = $(wildcard *.h*)
 HEADERS += $(wildcard demo/*.h*)
+
 BIN_DIR=build/bin
 OBJ_DIR=build/obj
 
@@ -76,7 +77,8 @@ GEN_SAMPLES_OUTPUT += $(patsubst samples/graph_%.dot,out/stdout_graph_%.txt,$(GE
 #GEN_SAMPLES_PLOT = $(patsubst samples/%.txt,out/sample_%.svg,$(GEN_SAMPLE_FILES))
 
 DOT_FILES=$(wildcard out/*.dot)
-SVG_FILES=$(patsubst out/%.dot,out/svg/%.svg,$(DOT_FILES))
+SVG_FILES=$(patsubst out/%.dot,out/svg/%_dot.svg,$(DOT_FILES))
+SVG_FILES+=$(patsubst out/%.dot,out/svg/%_neato.svg,$(DOT_FILES))
 
 # default target
 all: $(EXEC_FILES)
@@ -126,11 +128,17 @@ out/stdout_graph_%.txt: samples/graph_%.dot $(BIN_DIR)/read_graph makefile
 	@-$(BIN_DIR)/read_graph $< > $@;\
 	STATUS=$$?; echo "file $<: exit with $$STATUS" >> build/runsam.log
 
-out/svg/%.svg : out/%.dot
-	mkdir -p out/svg
-	echo "generating $@"
-	dot   -Tsvg -Nfontsize=24 $< >out/svg/$(notdir $(basename $@))_dot.svg
-	neato -Tsvg -Nfontsize=24 -Elen=1.5 $< >out/svg/$(notdir $(basename $@))_neato.svg
+
+create_svg:
+	mkdir -p out/svg/
+
+out/svg/%_dot.svg: out/%.dot create_svg
+	@echo "generating $@"
+	@dot   -Tsvg -Nfontsize=24 $< >out/svg/$(notdir $(basename $@))_dot.svg
+
+out/svg/%_neato.svg: out/%.dot create_svg
+	@echo "generating $@"
+	@neato -Tsvg -Nfontsize=24 -Elen=2.0 $< >out/svg/$(notdir $(basename $@))_neato.svg
 
 show: $(SRC_FILES)
 	@echo SRC_FILES=$(SRC_FILES)
@@ -151,6 +159,11 @@ doc:
 	doxygen misc/doxyfile 1>build/doxygen.stdout 2>build/doxygen.stderr
 	xdg-open build/html/index.html
 
+doc2:
+	make runsam -j4
+	make svg -j4
+	make doc
+
 clean:
 	@-rm $(OBJ_DIR)/*
 	@-rm $(BIN_DIR)/*
@@ -196,8 +209,8 @@ $(BIN_DIR)/%: $(OBJ_DIR)/%.o
 #	$(CXX) -o bin/test_catch $(OBJ_DIR)/test_catch.o -s
 #	@echo "done target $@"
 
-$(BIN_DIR)/test_catch:
-	$(CXX) -o $(BIN_DIR)/test_catch test_catch.cpp -s
+$(BIN_DIR)/test_catch: udgcd.hpp test_catch.cpp
+	$(CXX) -o $(BIN_DIR)/test_catch test_catch.cpp -s $(CFLAGS)
 	@echo "done target $@"
 
 # -s option: also shows successful test results

misc/doxyfile

@@ -803,7 +803,7 @@ WARN_LOGFILE           =
 # spaces. See also FILE_PATTERNS and EXTENSION_MAPPING
 # Note: If this tag is empty the current directory is searched.
 
-INPUT                  = ./ misc/
+INPUT                  = ./ misc/ demo/
 
 # This tag can be used to specify the character encoding of the source files
 # that doxygen parses. Internally doxygen uses the UTF-8 encoding. Doxygen uses
@@ -896,7 +896,7 @@ EXAMPLE_RECURSIVE      = NO
 # that contain images that are to be included in the documentation (see the
 # \image command).
 
-IMAGE_PATH             = out/
+IMAGE_PATH             = out/ misc/
 
 # The INPUT_FILTER tag can be used to specify a program that doxygen should
 # invoke to filter for each input file. Doxygen will invoke the filter program