Kill Graphpass on graphs with greater than 40,000 nodes (Issue #22) (#23

) * Upgrade gexf output to 1.3. * Kill graphpass on nodes > MAX_NODES (currently set to 40,000). * Review request for using CONSTANT instead of "40k" * Create verbose mode (-v or --verbose). * Set MAX_NODES to 50000. * "Less" should be "Fewer" in error message.
archivesunleashed · Jun 12, 2018 · d17b57c8d8eb7c7927fb7b5b74c748b293efba35 · d17b57c
1 parent 993c231
commit d17b57c8d8eb7c7927fb7b5b74c748b293efba35
Unified Split

Showing with 64 additions and 42 deletions.

+4 −1 headers/graphpass.h

+0 −2 src/analyze.c

+12 −9 src/filter.c

+7 −7 src/gexf.c

+23 −20 src/graphpass.c

+6 −3 src/io.c

+12 −0 src/quickrun.c
diff --git a/headers/graphpass.h b/headers/graphpass.h
@@ -30,6 +30,7 @@
 #include <getopt.h>

 typedef enum { false, true } bool;
+typedef enum { FAIL, WARN, COMM } broadcast;

 igraph_t g;
 igraph_attribute_table_t att;
@@ -44,6 +45,7 @@ bool ug_report; /**< Include a report? */
 bool ug_gformat; /**< Graph format - true is "GEXF" false is "GRAPHML" */
 bool ug_quickrun; /**< Lightweight visualization run */
 bool ug_save; /**< If false, does not save graphs at all (for reports) */
+bool ug_verbose; //**< Verbose mode (default off). */
 bool CALC_WEIGHTS;
 igraph_vector_t WEIGHTED; /**< If greater than 0, conducts weighted analysis */

@@ -65,14 +67,15 @@ igraph_vector_t WEIGHTED; /**< If greater than 0, conducts weighted analysis */
 #define VIZ_SET_LARGE "viz_large.h"
 #define VIZ_SET_SMALL "vis_small.h"

-/* Required External libraries */
+/* Default Settings */
 #define MAX_METHODS 9
 #define ALL_METHODS "abdehiopr"
 #define SIZE_DEFAULT "Degree"
 #define SIZE_DEFAULT_CHAR 'd'
 #define COLOR_BASE "WalkTrapModularity"
 #define PAGERANK_DAMPING 0.85 /**< chance random walk will not restart */
 #define LAYOUT_DEFAULT_CHAR 'f'
+#define MAX_NODES 50000 /**< number of nodes in graph before shut down */

 #define NELEMS(x)  (sizeof(x) / sizeof((x)[0]))

diff --git a/src/analyze.c b/src/analyze.c
@@ -207,7 +207,6 @@ int paired_t_stat (igraph_vector_t *v1, igraph_vector_t *v2, igraph_real_t *pv,
  igraph_real_t pvalue;
  
  for (long int i=0; i<igraph_vector_size(v2); i++) {
-    //printf("diff %f", igraph_vector_e(v1,i) - igraph_vector_e(v2,i));
    VECTOR(diff)[i] = igraph_vector_e(v1,i) - igraph_vector_e(v2,i);
  }
  *pv = t_test_vector(&diff, igraph_vector_size(&diff)-1);
@@ -428,7 +427,6 @@ int produceRank(igraph_vector_t *source, igraph_vector_t *v) {
    } else {
      VECTOR(rank_vals)[i] = i+1;
    }
-    //printf("I : %li, RANKVALS: %f", i, VECTOR(rank_vals)[i]);
  }
  for (long int i=0; i < source_size; i++) {
    for (long int j=0; j < source_size; j++) {
diff --git a/src/filter.c b/src/filter.c
@@ -168,7 +168,6 @@ int create_filtered_graph(igraph_t *graph, double cutoff, int cutsize, char* att
  VANV(&g2, "Outdegree", &odeg);
  // print size here.
  produceRank(&size, &rank);
-  printf("finishProduce Rank");
  SETVANV(&g2, "DegreeRank", &rank);
  set_size(&g2, &size, 100);
  centralization(&g2, "Betweenness");
@@ -294,21 +293,25 @@ int filter_graph() {
  float percentile;
  int cutsize;
  if (ug_quickrun == true) {
-    printf("\n\nQuickrun requested.\n\n");
-    printf("Quickrun does no filtering, and provides layout information\n");
-    printf("based on Degree (nodesize), Walktrap Modularity (color), and\n");
-    printf("the Fructerman-Rheingold algorithm to maximize space between\n");
-    printf("nodes.\n\n");
-    printf("Quickrun is quicker, but less informative in terms of output.\n");
+    if (ug_verbose == true) {
+      printf("\n\nQuickrun requested.\n\n");
+      printf("Quickrun does no filtering, and provides layout information\n");
+      printf("based on Degree (nodesize), Walktrap Modularity (color), and\n");
+      printf("the Fructerman-Rheingold algorithm to maximize space between\n");
+      printf("nodes.\n\n");
+      printf("Quickrun is quicker, but less informative in terms of output.\n");
+    }
    quickrunGraph();
    igraph_destroy(&g);
    return(0);
  }
  /* if (CALC_WEIGHTS == false) {igraph_vector_init(&WEIGHTED, NODESIZE);}*/
  percentile = (ug_percent > 0.99) ? fix_percentile() : ug_percent;
  cutsize = round((double)NODESIZE * percentile);
-  printf("Filtering the graphs by %f will reduce the graph size by %d \n", ug_percent, cutsize);
-  printf("This will produce a graph with %d nodes.\n", (NODESIZE - cutsize));
+  if (ug_verbose == true) {
+    printf("Filtering the graphs by %f will reduce the graph size by %d \n", ug_percent, cutsize);
+    printf("This will produce a graph with %d nodes.\n", (NODESIZE - cutsize));
+  }
  SETVANV(&g, "idRef", &idRef);
  analysis_all(&g);
  runFilters(&g, cutsize);
diff --git a/src/gexf.c b/src/gexf.c
@@ -73,24 +73,24 @@ extern int igraph_write_graph_gexf(const igraph_t *graph, FILE *outstream,
  const char *vprefix= prefixattr ? "v_" : "";
  const char *eprefix= prefixattr ? "e_" : "";
  
-  ret=fprintf(outstream, "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n");
+  ret=fprintf(outstream, "<?xml version=\"1.1\" encoding=\"UTF-8\"?>\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
-  ret=fprintf(outstream, "<gexf xmlns=\"http://www.gexf.net/1.2draft\"\n");
+  ret=fprintf(outstream, "<gexf xmlns=\"http://www.gexf.net/1.3draft\"\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
-  ret=fprintf(outstream, "         xmlns:viz=\"http://www.gexf.net/1.2draft/viz\"\n");
+  ret=fprintf(outstream, "         xmlns:viz=\"http://www.gexf.net/1.3draft/viz\"\n");
  ret=fprintf(outstream, "         xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\"\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
-  ret=fprintf(outstream, "         xsi:schemaLocation=\"http://www.gexf.net/1.2draft\n");
+  ret=fprintf(outstream, "         xsi:schemaLocation=\"http://www.gexf.net/1.3draft\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
-  ret=fprintf(outstream, "         http://www.gexf.net/1.2draft/gexf.xsd\"\n");
+  ret=fprintf(outstream, "         http://www.gexf.net/1.3draft/gexf.xsd\"\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
-  ret=fprintf(outstream, "         version=\"1.2\">\n");
+  ret=fprintf(outstream, "         version=\"1.3\">\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
  ret=fprintf(outstream, "<!-- Created by igraph -->\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
  ret=fprintf(outstream, "<meta lastmodifieddate=\"%.19s\">\n", ctime(&t));
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
-  ret=fprintf(outstream, "<creator>Borg-reducer filtering using Igraph</creator>\n");
+  ret=fprintf(outstream, "<creator>Graphpass filtering using Igraph by Archives Unleashed</creator>\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
  ret=fprintf(outstream, "<description> A Filtered Derivative Graph</description>\n");
  if (ret<0) IGRAPH_ERROR("Write failed", IGRAPH_EFILE);
diff --git a/src/graphpass.c b/src/graphpass.c
@@ -32,6 +32,7 @@


 #include "igraph.h"
+#include <stdlib.h>
 #include "graphpass.h"

 char* FILEPATH; /**< The filepath (ug_DIRECTORY + ug_FILENAME) */
@@ -44,8 +45,8 @@ bool ug_gformat = false;
 bool ug_report = false;
 /** Provide a quickrun with simple sizing, positioning and coloring **/
 bool ug_quickrun = false;
-
-int verbose_flag;
+/** Print out helper messages **/
+bool ug_verbose = false;

 int main (int argc, char *argv[]) {
  /* Experiments here 
@@ -57,11 +58,9 @@ int main (int argc, char *argv[]) {
    {
      static struct option long_options[] =
        {
-          /* These options set a flag. */
-          {"verbose", no_argument,       &verbose_flag, 1},
-          {"brief",   no_argument,       &verbose_flag, 0},
          /* These options don’t set a flag.
             We distinguish them by their indices. */
+          {"verbose", no_argument,       0, 'v'},
          {"no-save", no_argument,       0, 'n'},
          {"report",  no_argument,       0, 'r'},
          {"gexf",    no_argument,       0, 'g'},
@@ -76,7 +75,7 @@ int main (int argc, char *argv[]) {
        };
      /* getopt_long stores the option index here. */
      int option_index = 0;
-      c = getopt_long (argc, argv, "nrgqwf:p:m:o:d",
+      c = getopt_long (argc, argv, "vnrgqwf:p:m:o:d",
                       long_options, &option_index);

      /* Detect the end of the options. */
@@ -88,6 +87,9 @@ int main (int argc, char *argv[]) {
          /* If this option set a flag, do nothing else now. */
          if (long_options[option_index].flag != 0)
            break;
+        case 'v':
+          ug_verbose = true;
+          break;
        case 'n':
          ug_save = !ug_save;
          break;
@@ -127,12 +129,6 @@ int main (int argc, char *argv[]) {
        }
    }

-  /* Instead of reporting ‘--verbose’
-     and ‘--brief’ as they are encountered,
-     we report the final status resulting from them. */
-  if (verbose_flag)
-    puts ("verbose flag is set");
-
  /* Print any remaining command line arguments (not options). */
  if (optind < argc)
    {
@@ -154,12 +150,13 @@ int main (int argc, char *argv[]) {
  strncpy(path, ug_DIRECTORY, strlen(ug_DIRECTORY)+1);
  
  /** start output description **/
-  printf(">>>>>>>  GRAPHPASSING >>>>>>>> \n");
-  printf("DIRECTORY: %s \nSTRLEN PATH: %li \n", ug_DIRECTORY, strlen(path));
-  printf("OUTPUT DIRECTORY: %s\nPERCENTAGE: %f\n", ug_OUTPUT, ug_percent);
-  printf("FILE: %s\nMETHODS STRING: %s\n", ug_FILENAME, ug_methods);
-  printf("QUICKRUN: %i\nREPORT: %i\nSAVE: %i\n", ug_quickrun, ug_report, ug_save);
-  
+  if (ug_verbose == true) {
+    printf(">>>>>>>  GRAPHPASSING >>>>>>>> \n");
+    printf("DIRECTORY: %s \nSTRLEN PATH: %li \n", ug_DIRECTORY, strlen(path));
+    printf("OUTPUT DIRECTORY: %s\nPERCENTAGE: %f\n", ug_OUTPUT, ug_percent);
+    printf("FILE: %s\nMETHODS STRING: %s\n", ug_FILENAME, ug_methods);
+    printf("QUICKRUN: %i\nREPORT: %i\nSAVE: %i\n", ug_quickrun, ug_report, ug_save);
+  }
  /** try to be nice if user leaves out a '/' **/
  if (ug_FILENAME[0] == '/' && ug_DIRECTORY[strlen(ug_DIRECTORY)] == '/' ){
    path[strlen(path)+1] = '\0';  // remove end slash
@@ -174,10 +171,16 @@ int main (int argc, char *argv[]) {
  int filepathsize = sizeOfPath + sizeOfFile;
  FILEPATH = malloc(filepathsize + 1);
  snprintf(FILEPATH, filepathsize, "%s%s", path, ug_FILENAME);
-  printf("Running graphpass on file: %s\n", FILEPATH);
-  
+  if (ug_verbose == true) {
+    printf("Running graphpass on file: %s\n", FILEPATH);
+  }
  load_graph(FILEPATH);
  free(FILEPATH);
+  if (igraph_vcount(&g) > MAX_NODES) {
+    printf ("FAIL >>> Graphpass can only conduct analysis on graphs with fewer than %i nodes.\n", MAX_NODES);
+    printf ("FAIL >>> Exiting...\n");
+    exit(EXIT_FAILURE);
+  }
  
  /** start the filtering based on values and methods **/
  filter_graph();
diff --git a/src/io.c b/src/io.c
@@ -46,7 +46,9 @@ extern int load_graph (char* filename) {
  }
  igraph_read_graph_graphml(&g, fp, 0);
  NODESIZE = igraph_vcount(&g);
-  printf("Successfully ingested graph with %li nodes.\n", (long int)NODESIZE);
+  if (ug_verbose) {
+    printf("Successfully ingested graph with %li nodes.\n", (long int)NODESIZE);
+  }
  fclose(fp);
  return (0);
 }
@@ -87,7 +89,9 @@ extern int write_graph(igraph_t *graph, char *attr) {
    strncat(path, ".graphml", 8);
  }
  if (ug_save == true) {
-    printf("Writing output to: %s\n", path);
+    if (ug_verbose == true) {
+      printf("Writing output to: %s\n", path);
+    }
    fp = fopen(path, "w");
    if (ug_gformat) {
      igraph_write_graph_gexf(graph, fp, 1);
@@ -96,7 +100,6 @@ extern int write_graph(igraph_t *graph, char *attr) {
    }
    fclose(fp);
  }
-  
  return 0;
 }

diff --git a/src/quickrun.c b/src/quickrun.c
@@ -30,11 +30,23 @@
 extern int quickrunGraph() {
  igraph_vector_t size;
  igraph_vector_init(&size, igraph_vcount(&g));
+  if (ug_verbose == true)
+    printf ("Calculating Degree... \n");
  calc_degree(&g, 'd');
+  if (ug_verbose == true)
+    printf ("Calculating Modularity... \n");
  calc_modularity(&g);
+  if (ug_verbose == true)
+    printf ("Assigning Colors... \n");
  colors(&g);
+  if (ug_verbose == true)
+    printf("Assigning Degree values... \n");
  VANV(&g, "Degree", &size);
+  if (ug_verbose == true)
+    printf("Scaling graph with degree values... \n");
  set_size(&g, &size, 100);
+  if (ug_verbose == true)
+    printf("Producing layout details... \n");
  layout_graph(&g, 'f');
  write_graph(&g, "-");
  igraph_vector_destroy(&size);