diff --git a/Makefile b/Makefile
index 492fddf376b92b93390529f5de460efabc5d356d..4c47334a9289017e992095e1b082cb132ab9d70e 100644
--- a/Makefile
+++ b/Makefile
@@ -21,7 +21,7 @@ TP1: TP1.o file.o
TP2: TP2.o struct_tp2.o
$(CC) $(OPTIONS) -o $@ $^ $(MATH) -g
-TP3: TP3.o plotout.o
+TP3: TP3.o plotout.o file.o
$(CC) $(OPTIONS) -o $@ $^ $(MATH) -g
TP4: TP4.o
diff --git a/TP1 b/TP1
index b6d00fefb1ec8b3b30bb4af7a13066bb9eac6402..122f217d6fcaf97f4b177fd9e7e99deaf2116a7f 100755
Binary files a/TP1 and b/TP1 differ
diff --git a/TP2 b/TP2
index 8c90ef31e5334fcce47f26e929f863d6d82585a1..9099ac9b2dbc979d06b03ab0134fdc79ab8fa6fe 100755
Binary files a/TP2 and b/TP2 differ
diff --git a/TP3 b/TP3
index 2c865c5723dcb9cea539ca0d5767943ff098fea6..98db9dcd6f1a44675eb4af731a5a225a19aa83c8 100755
Binary files a/TP3 and b/TP3 differ
diff --git a/TP3.c b/TP3.c
index 7f94b4aac53952a0c2e475b568db57edb2511d30..ad59b6eb42d63fce9a215d498e3906f293843b42 100644
--- a/TP3.c
+++ b/TP3.c
@@ -1,5 +1,6 @@
#include "./implems.c"
#include "./plotout.h"
+#include "./file.h"
#include <stdio.h>
#include <stdlib.h>
#include <limits.h>
@@ -109,6 +110,256 @@ int pagerank(edgelist *g, long double *distr_res, long double alpha, unsigned lo
}
+// fonction réalisant nb_iter itérations de personalized power pagerank
+// distr_res doit être malloc de taille g->
+// p0 doit être un vecteur de probabilités de taille g->n dont la somme est 1
+int pers_pagerank(edgelist *g, long double *distr_res, long double *p0, long double alpha, unsigned long nb_iter) {
+ unsigned long i, j;
+ unsigned long *out_deg;
+ long double *prob_t = distr_res;
+ long double *prob_t1;
+ long double *swap_pointeur;
+ out_deg = (unsigned long*) malloc((g->n)*sizeof(unsigned long));
+ prob_t1 = (long double*) malloc((g->n)*sizeof(long double));
+
+
+
+ // calcul les degrés sortants
+ calc_out_deg(g, out_deg);
+
+
+ // initialisation des probas
+ for (i = 0; i < g->n; i += 1) {
+ prob_t[i] = 1./((long double) (g->n));
+ }
+
+
+ // itérations de PageRank
+ for (i = 0; i < nb_iter; i += 1) {
+ // mise à zero de prob_t1
+ for (j = 0; j < g->n; j += 1) {
+ prob_t1[j] = 0;
+ }
+
+ // itération sur les arêtes pour chaque quantité
+ for (j = 0; j < g->e; j += 1) {
+ prob_t1[g->edges[j].t] += (1 - alpha)*(prob_t[g->edges[j].s]/out_deg[g->edges[j].s]);
+ }
+
+ // ajout du poids alpha en plus (pondéré par P0) /!\ remise à n pour le poid de P0
+ for (j = 0; j < g->n; j +=1) {
+ prob_t1[j] += alpha*p0[j]; // renvoie une proportion alpha sur p0
+ // prob_t1[j] += alpha*(1./((long double) (g->n))); // renvoie une proportion alpha dans le reste du tableau
+ }
+
+ // renormalisation du vecteur (avec poid pondéré par P0) /!\ remise à n pour le poid de p0
+ long double v = 0;
+ for (j = 0; j < g->n; j += 1) {
+ v += prob_t1[j];
+ }
+ long double vprim = (1-v)/((long double) (g->n));
+ v = (1-v);
+ for (j = 0; j < g->n; j += 1) {
+ // prob_t1[j] += v*p0[j]; // normalise uniquement sur les composantes de P0
+ prob_t1[j] += vprim; // normalise sur la totalité du vecteur
+ }
+ // échange des pointeurs pour prochaine itération
+ swap_pointeur = prob_t;
+ prob_t = prob_t1;
+ prob_t1 = swap_pointeur;
+ }
+ // si nombre impair d'itération, copie du résultat final dans le bon tableau
+ if (nb_iter%2 == 1) {
+ for (i = 0; i < g->n; i += 1) {
+ distr_res[i] = prob_t[i];
+ }
+ free(prob_t);
+ }
+ else {
+ free(prob_t1);
+ }
+ free(out_deg);
+ return 0;
+}
+
+
+
+// fonction déterminant quels noeuds sont dans la catégorie donnée
+// requires
+ // node_res est un tableau de taille g->n
+ // filepath est le chemin vers le fichier d'association noeud ID -> category ID
+// ensures
+ // node_res contient 1 si le noeud est dans la catégorie demandée, 0 sinon
+
+int mark_in_cat(edgelist *g, unsigned long *node_res, char *filepath, char *filecatdir, unsigned long cat_wanted) {
+ // parcour en largeur pour trouver les sous catégories de échec
+ adjlist *cat_graph;
+ cat_graph = al_readedgelist(filecatdir);
+ mkadjlist(cat_graph);
+
+ // fait un BFS
+ // ################################
+ file_t F = new_file(cat_graph->n);
+ int *marked = (int *)malloc((cat_graph->n) * sizeof(int));
+ for (unsigned long i = 0; i < cat_graph->n; i++)
+ {
+ marked[i] = -1;
+ }
+ enfile(F, cat_wanted);
+ marked[cat_wanted] = 0;
+ unsigned long tmp = cat_wanted;
+ while (!is_empty(F))
+ {
+ tmp = defile(F);
+ // printf("%li %li %li %li\n", tmp, marked[tmp], cat_graph->cd[tmp], cat_graph->cd[tmp+1]);
+ for (unsigned long i = cat_graph->cd[tmp]; i < cat_graph->cd[tmp + 1]; i++)
+ {
+ if (marked[cat_graph->adj[i]] == -1)
+ {
+ if (enfile(F, cat_graph->adj[i]) == 1)
+ {
+ marked[cat_graph->adj[i]] = marked[tmp] + 1;
+ }
+ else
+ {
+ printf("\n\nPile trop petite !\n\n");
+ }
+ }
+ }
+ }
+ unsigned long no_marque = 0;
+ for (unsigned long i = 0; i < cat_graph->n; i++)
+ {
+ if (marked[i] == -1 && cat_graph->cd[0] != 0)
+ {
+ no_marque++;
+ }
+ }
+ // ################################
+
+
+
+ // trouve les noeuds qui ont été marqués
+ int count_cat = 0;
+ for (unsigned long i = 0; i < cat_graph->n; i += 1) {
+ if (marked[i] != -1) {
+ count_cat += 1;
+ }
+ }
+
+ printf("cat : %lu related nb : %i/%lu\n", cat_wanted, count_cat, cat_graph->n);
+
+
+
+ free_adjlist(cat_graph);
+ // ---------------------------------------------------------------------
+ // --------------------------------marked != -1 si subcat de wanted_cat---
+ // ---------------------------------------------------------------------
+
+
+
+
+ // mise à zero du tableau resultat
+ for (unsigned long i = 0; i < g->n; i += 1) {
+ node_res[i] = 0;
+ }
+
+ unsigned long node_ID, cat_ID;
+ char test_end;
+
+ // lis le fichier d'entrée
+ FILE *file = fopen(filepath, "r");
+ char line[1000];
+ // ignore les premières lignes de commentaire
+ while(fgets(line, sizeof line, file) && *line == '#') {}
+
+ // lit les entrées
+ while(fscanf(file, "%lu%c", &node_ID, &test_end) == 2) {
+ // printf("node start : %lu\n", node_ID);
+ if (test_end != '\n') {
+ while(fscanf(file, "%lu%c", &cat_ID, &test_end)==2) {
+ // printf("lié à %lu\n", cat_ID);
+ if (marked[cat_ID] != -1) {
+ // printf("ouais !\n");
+ node_res[node_ID] = 1;
+ }
+ if (test_end == '\n') {break;}
+ }
+ }
+ }
+
+
+
+ free(marked);
+ return 0;
+}
+
+
+// trouve les subcatégories de la catégorie passée en paramètre
+int find_sub_cat(char *filecatdir, unsigned long cat_wanted) {
+ adjlist *cat_graph;
+ cat_graph = al_readedgelist(filecatdir);
+ mkadjlist(cat_graph);
+
+ // fait un BFS
+ // ################################
+ file_t F = new_file(cat_graph->n);
+ int *marked = (int *)malloc((cat_graph->n) * sizeof(int));
+ for (unsigned long i = 0; i < cat_graph->n; i++)
+ {
+ marked[i] = -1;
+ }
+ enfile(F, cat_wanted);
+ marked[cat_wanted] = 0;
+ unsigned long tmp = cat_wanted;
+ while (!is_empty(F))
+ {
+ tmp = defile(F);
+ // printf("%li %li %li %li\n", tmp, marked[tmp], cat_graph->cd[tmp], cat_graph->cd[tmp+1]);
+ for (unsigned long i = cat_graph->cd[tmp]; i < cat_graph->cd[tmp + 1]; i++)
+ {
+ if (marked[cat_graph->adj[i]] == -1)
+ {
+ if (enfile(F, cat_graph->adj[i]) == 1)
+ {
+ marked[cat_graph->adj[i]] = marked[tmp] + 1;
+ }
+ else
+ {
+ printf("\n\nPile trop petite !\n\n");
+ }
+ }
+ }
+ }
+ unsigned long no_marque = 0;
+ for (unsigned long i = 0; i < cat_graph->n; i++)
+ {
+ if (marked[i] == -1 && cat_graph->cd[0] != 0)
+ {
+ no_marque++;
+ }
+ }
+ // ################################
+
+
+
+ // trouve les noeuds qui ont été marqués
+ int count_cat = 0;
+ for (unsigned long i = 0; i < cat_graph->n; i += 1) {
+ if (marked[i] != -1) {
+ count_cat += 1;
+ }
+ }
+
+ printf("cat : %lu related nb : %i/%lu\n", cat_wanted, count_cat, cat_graph->n);
+
+
+
+ free(marked);
+ free_adjlist(cat_graph);
+ return 1;
+}
+
void debug_print_tabl(long double *tab) {
@@ -214,75 +465,157 @@ int main(int argc, char **argv) {
// printf("- Overall time = %I64dh%I64dm%I64ds\n",(t2-t1)/3600,((t2-t1)%3600)/60,((t2-t1)%60));
+
+
+
+
+
// fonctions du TP3
+ // long double *p_distrib;
+ // unsigned long *highest;
+ // unsigned long *lowest;
+ // p_distrib = (long double*) malloc((g->n)*sizeof(long double));
+ // highest = (unsigned long*) malloc(5*sizeof(unsigned long));
+ // lowest = (unsigned long*) malloc(5*sizeof(unsigned long));
+
+
+ // pagerank(g, p_distrib, 0.15, 70);
+
+
+ // find_5(g, p_distrib, lowest, highest);
+
+
+
+
+
+ // //TABLEAUX pour les CORRELATIONS
+
+ // long double *p_15 = p_distrib;
+ // long double *p_1;
+ // long double *p_2;
+ // long double *p_5;
+ // long double *p_9;
+ // unsigned long *out_deg;
+ // unsigned long *in_deg;
+
+ // p_1 = (long double*) malloc((g->n)*sizeof(long double));
+ // p_2 = (long double*) malloc((g->n)*sizeof(long double));
+ // p_5 = (long double*) malloc((g->n)*sizeof(long double));
+ // p_9 = (long double*) malloc((g->n)*sizeof(long double));
+ // out_deg = (unsigned long*) malloc((g->n)*sizeof(unsigned long));
+ // in_deg = (unsigned long*) malloc((g->n)*sizeof(unsigned long));
+
+ // printf("out_deg processing...\n");
+ // calc_out_deg(g, out_deg);
+ // printf("in_deg processing...\n");
+ // calc_in_deg(g, in_deg);
+ // printf("pagerank 0.1 processing...\n");
+ // pagerank(g, p_1, 0.1, 70);
+ // printf("pagerank 0.2 processing...\n");
+ // pagerank(g, p_2, 0.2, 70);
+ // printf("pagerank 0.5 processing...\n");
+ // pagerank(g, p_5, 0.5, 70);
+ // printf("pagerank 0.9 processing...\n");
+ // pagerank(g, p_9, 0.9, 70);
+
+
+ // printf("output plots ...\n");
+ // plot_out_2D_ld_ul("1.csv", g->n, p_15, in_deg);
+ // plot_out_2D_ld_ul("2.csv", g->n, p_15, out_deg);
+ // plot_out_2D_ld_ld("3.csv", g->n, p_15, p_1);
+ // plot_out_2D_ld_ld("4.csv", g->n, p_15, p_2);
+ // plot_out_2D_ld_ld("5.csv", g->n, p_15, p_5);
+ // plot_out_2D_ld_ld("6.csv", g->n, p_15, p_9);
+
+
+ // free(p_1);
+ // free(p_2);
+ // free(p_5);
+ // free(p_9);
+ // free(p_15);
+
+
+
+
+
+
+
+
+
+ // personalized PAGERANK
+
+ // ################# DATA ##############
+ // ID of category chess : 691713
+ // ID of page Magnus carlsen : 442682
+ // ID of category boxing : 738624
+ // #####################################
+
+
long double *p_distrib;
- unsigned long *highest;
- unsigned long *lowest;
+ long double *p0_magnus;
+ unsigned long *nodes;
+ unsigned long *nodes_2;
p_distrib = (long double*) malloc((g->n)*sizeof(long double));
- highest = (unsigned long*) malloc(5*sizeof(unsigned long));
- lowest = (unsigned long*) malloc(5*sizeof(unsigned long));
+ nodes = (unsigned long*) malloc((g->n)*sizeof(unsigned long));
+ nodes_2 = (unsigned long*) malloc((g->n)*sizeof(unsigned long));
+
+ printf("finding nodes related to chess category...\n");
+ mark_in_cat(g, nodes, "../alr21--pageCategList--enwiki--20071018.txt", "../alr21--categDAG--dirLinks--enwiki-20071018.txt", 691713);
+
+
- pagerank(g, p_distrib, 0.15, 70);
- find_5(g, p_distrib, lowest, highest);
+ // ROOTED PAGERANK of MAGNUS CARLSEN
+ p0_magnus = (long double*) malloc((g->n)*sizeof(long double));
+ for (unsigned long i = 0; i < g->n; i += 1) {
+ p0_magnus[i] = 0;
+ }
+ p0_magnus[442682] = 1; //page magnus carlsen
+ printf("computing rooted pagerank of Magnus Carlsen...\n");
+ pers_pagerank(g, p_distrib, p0_magnus, 0.55, 50);
+ plot_out_2D_ld_ul("magnus.csv", g->n, p_distrib, nodes);
- // TABLEAUX pour les CORRELATIONS
- long double *p_15 = p_distrib;
- long double *p_1;
- long double *p_2;
- long double *p_5;
- long double *p_9;
- unsigned long *out_deg;
- unsigned long *in_deg;
- p_1 = (long double*) malloc((g->n)*sizeof(long double));
- p_2 = (long double*) malloc((g->n)*sizeof(long double));
- p_5 = (long double*) malloc((g->n)*sizeof(long double));
- p_9 = (long double*) malloc((g->n)*sizeof(long double));
- out_deg = (unsigned long*) malloc((g->n)*sizeof(unsigned long));
- in_deg = (unsigned long*) malloc((g->n)*sizeof(unsigned long));
- printf("out_deg processing...\n");
- calc_out_deg(g, out_deg);
- printf("in_deg processing...\n");
- calc_in_deg(g, in_deg);
- printf("pagerank 0.1 processing...\n");
- pagerank(g, p_1, 0.1, 70);
- printf("pagerank 0.2 processing...\n");
- pagerank(g, p_2, 0.2, 70);
- printf("pagerank 0.5 processing...\n");
- pagerank(g, p_5, 0.5, 70);
- printf("pagerank 0.9 processing...\n");
- pagerank(g, p_9, 0.9, 70);
-
-
- printf("output plots ...\n");
- plot_out_2D_ld_ul("1.csv", g->n, p_15, in_deg);
- plot_out_2D_ld_ul("2.csv", g->n, p_15, out_deg);
- plot_out_2D_ld_ld("3.csv", g->n, p_15, p_1);
- plot_out_2D_ld_ld("4.csv", g->n, p_15, p_2);
- plot_out_2D_ld_ld("5.csv", g->n, p_15, p_5);
- plot_out_2D_ld_ld("6.csv", g->n, p_15, p_9);
-
- // long double last = p_distrib[0];
- // for(unsigned long i =0; i < g->n; i +=1) {
- // if (last != p_distrib[i]) {
- // printf("|%Le|\n", p_distrib[i]);
- // }
- // }
-
- free(p_1);
- free(p_2);
- free(p_5);
- free(p_9);
- free(p_15);
+
+
+
+ // restart vector for chess && boxing
+ printf("finding nodes related to boxing category...\n");
+ mark_in_cat(g, nodes_2, "../alr21--pageCategList--enwiki--20071018.txt", "../alr21--categDAG--dirLinks--enwiki-20071018.txt", 738624);
+
+ printf("construct of the chess/boxing restart vector ...\n");
+ long double *p0_chess_box = p0_magnus;
+ unsigned long count = 0;
+ for (unsigned long i = 0; i < g->n; i += 1) {
+ count += nodes[i] + nodes_2[i];
+ p0_chess_box[i] = nodes[i] + nodes_2[i];
+ }
+ for (unsigned long i = 0; i < g->n; i += 1) {
+ p0_chess_box[i] = p0_chess_box[i]/count;
+ }
+
+ printf("computing personalized pagerank of chess and box...\n");
+ pers_pagerank(g, p_distrib, p0_chess_box, 0.15, 50);
+
+ for (unsigned long i = 0; i < g->n; i += 1) {
+ if (nodes[i] == 1 || nodes_2[i] == 1) {
+ nodes[i] = 1;
+ }
+ }
+
+ plot_out_2D_ld_ul("chessbox.csv", g->n, p_distrib, nodes);
+
+
+ free(nodes);
+ free(p_distrib);
return 0;
}
\ No newline at end of file
diff --git a/chessbox.csv b/chessbox.csv
new file mode 100644
index 0000000000000000000000000000000000000000..78c1fd5ee7b05b55ae83614f490467c289a24e77
Binary files /dev/null and b/chessbox.csv differ
diff --git a/magnus.csv b/magnus.csv
new file mode 100644
index 0000000000000000000000000000000000000000..6035dffd4d9fd6603e795d0b9c73bc061f8bbe28
Binary files /dev/null and b/magnus.csv differ
diff --git a/plotout.h b/plotout.h
index 1df89a41b32d4b1b72611ece2e71ce09ee0cf634..252aaea6f180ae34831b08707983da3613704404 100644
--- a/plotout.h
+++ b/plotout.h
@@ -10,5 +10,4 @@ int plot_out_2D_ld_ul(char* name, unsigned long size, long double *x, unsigned l
int plot_out_2D_ld_ld(char* name, unsigned long size, long double *x, long double *y);
-
#endif
\ No newline at end of file
diff --git a/trace_graphe.py b/trace_graphe.py
index ab82dab241717d8a93a330d3cef18f5891bf82aa..cebdcb324995a021787f621e082b4b17f07e60b2 100644
--- a/trace_graphe.py
+++ b/trace_graphe.py
@@ -21,21 +21,18 @@ data = pd.read_csv("../plots/1.csv", sep=",")
#plot des données
-plt.scatter(data['x'], data['y'])
+#plt.scatter(data['x'], data['y'])
+plt.plot(data['x'], data['y'], 'o')
#plot de la régression linéaire
m, b = np.polyfit(data['x'], data['y'], 1)
-plt.plot(data['x'], m*data['x']+b, color = "red")
-
-
+plt.plot(data['x'], m*data['x']+b, 'r.')
-plt.xlim(0,0.0025)
-plt.ylim(0,250000)
+plt.xscale("log")
+plt.yscale("log")
+#plt.xlim(0, 0.0025)
+plt.ylim(1, 200000)
-
-
-plt.xscale("linear")
-plt.yscale("linear")
#valeurs d'échelle : {"linear", "log", "symlog", "logit", ...}
plt.title("degré entrant en fonction de pagerank")
@@ -43,6 +40,9 @@ plt.xlabel("pagerank (α = 0.15)")
plt.ylabel("degré entrant")
+plt.savefig("1.png", dpi=100, format = 'png')
+
+
plt.show()
@@ -52,13 +52,26 @@ plt.show()
# PLOT 2
##########################
data = pd.read_csv("../plots/2.csv", sep=",")
-plt.scatter(data['x'], data['y'])
+
+
+#plt.scatter(data['x'], data['y'])
+plt.plot(data['x'], data['y'], 'o')
+
+#plot de la régression linéaire
+m, b = np.polyfit(data['x'], data['y'], 1)
+plt.plot(data['x'], m*data['x']+b, 'r.')
+
+plt.xscale("log")
+plt.yscale("log")
plt.title("degré sortant en fonction de pagerank")
plt.xlabel("pagerank (α = 0.15)")
plt.ylabel("degré sortant")
-plt.xlim(0,0.0035)
+#plt.xlim(0,0.0020)
+
+
+plt.savefig("2.png", dpi=100, format = 'png')
@@ -69,15 +82,24 @@ plt.show()
# PLOT 3
##########################
data = pd.read_csv("../plots/3.csv", sep=",")
-plt.scatter(data['x'], data['y'])
+
+
+plt.plot(data['x'], data['y'], 'o')
+
+#plot de la régression linéaire
+m, b = np.polyfit(data['x'], data['y'], 1)
+plt.plot(data['x'], m*data['x']+b, 'r.', markersize=10)
+
+plt.xscale("log")
+plt.yscale("log")
plt.title("pagerank (α = 0.15) vs pagerank (α = 0.1)")
plt.xlabel("pagerank (α = 0.15)")
plt.ylabel("pagerank (α = 0.1)")
-plt.xlim(0,0.0035)
-plt.ylim(0,0.005)
+plt.savefig("3.png", dpi=100, format = 'png')
+
plt.show()
@@ -86,15 +108,24 @@ plt.show()
# PLOT 4
##########################
data = pd.read_csv("../plots/4.csv", sep=",")
+
+
plt.scatter(data['x'], data['y'])
+#plot de la régression linéaire
+m, b = np.polyfit(data['x'], data['y'], 1)
+#plt.plot(data['x'], m*data['x']+b, 'r.')
+
+plt.xscale("log")
+plt.yscale("log")
+
plt.title("pagerank (α = 0.15) vs pagerank (α = 0.2)")
plt.xlabel("pagerank (α = 0.15)")
plt.ylabel("pagerank (α = 0.2)")
-plt.xlim(0,0.0035)
-plt.ylim(0,0.005)
+plt.savefig("4.png", dpi=100, format = 'png')
+
plt.show()
@@ -103,13 +134,22 @@ plt.show()
# PLOT 5
##########################
data = pd.read_csv("../plots/5.csv", sep=",")
+
+
plt.scatter(data['x'], data['y'])
+#plot de la régression linéaire
+m, b = np.polyfit(data['x'], data['y'], 1)
+#plt.plot(data['x'], m*data['x']+b, 'r.')
+
+plt.xscale("log")
+plt.yscale("log")
+
plt.title("pagerank (α = 0.15) vs pagerank (α = 0.5)")
plt.xlabel("pagerank (α = 0.15)")
plt.ylabel("pagerank (α = 0.5)")
-plt.xlim(0,0.0035)
-plt.ylim(0,0.0025)
+
+plt.savefig("5.png", dpi=100, format = 'png')
@@ -120,14 +160,127 @@ plt.show()
# PLOT 6
##########################
data = pd.read_csv("../plots/6.csv", sep=",")
+
+
plt.scatter(data['x'], data['y'])
+
+#plot de la régression linéaire
+m, b = np.polyfit(data['x'], data['y'], 1)
+#plt.plot(data['x'], m*data['x']+b, 'r.')
+
+plt.xscale("log")
+plt.yscale("log")
+
+
plt.title("pagerank (α = 0.15) vs pagerank (α = 0.9)")
plt.xlabel("pagerank (α = 0.15)")
plt.ylabel("pagerank (α = 0.9)")
-plt.xlim(0,0.0035)
-plt.ylim(0,0.0005)
+
+plt.savefig("6.png", dpi=100, format = 'png')
+
+plt.show()
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+##########################
+# MAGNUS plot
+##########################
+
+
+data_magnus = pd.read_csv("./magnus.csv", sep = ",")
+work = data_magnus.sort_values(by = 'x', ascending = False)
+ord_pagerank = list(work['x'])
+ord_incat = list(work['y'])
+
+fig, axs = plt.subplots(2, sharex=True)
+fig.suptitle('Corrélation entre la proximité aux noeuds et les catégories')
+axs[0].plot(range(len(ord_pagerank)), ord_pagerank)
+
+axs[0].set_xscale("log")
+axs[0].set_yscale("log")
+
+
+summed_incat = []
+sum = 0
+for i in range(len(ord_incat)):
+ sum += ord_incat[i]
+ summed_incat.append(sum)
+
+
+
+
+
+axs[1].plot(range(len(summed_incat)), summed_incat)
+axs[1].set_xscale("log")
+axs[1].set_yscale("linear")
+axs[0].axvline(x = 2000000, color = "black", linestyle='--')
+axs[1].axvline(x = 2000000, color = "black", linestyle='--')
+
+
+axs[1].set_xlabel("Rang du noeud")
+axs[0].set_ylabel("Proximité au noeud \"magnus carlsen\"")
+axs[1].set_ylabel("Nombre de noeuds dans les sous-catégories de \"échec\"")
+
+
+plt.show()
+
+
+
+
+
+##########################
+# CHESS & BOX plot
+##########################
+
+
+data_magnus = pd.read_csv("./chessbox.csv", sep = ",")
+work = data_magnus.sort_values(by = 'x', ascending = False)
+ord_pagerank = list(work['x'])
+ord_incat = list(work['y'])
+
+fig, axs = plt.subplots(2, sharex=True)
+fig.suptitle('Personalized pagerank avec un vecteur p0 orienté échec ou box')
+axs[0].plot(range(len(ord_pagerank)), ord_pagerank)
+
+axs[0].set_xscale("log")
+axs[0].set_yscale("log")
+
+
+summed_incat = []
+sum = 0
+for i in range(len(ord_incat)):
+ sum += ord_incat[i]
+ summed_incat.append(sum)
+
+
+
+
+
+axs[1].plot(range(len(summed_incat)), summed_incat)
+axs[1].set_xscale("log")
+axs[1].set_yscale("linear")
+axs[0].axvline(x = 2000000, color = "black", linestyle='--')
+axs[1].axvline(x = 2000000, color = "black", linestyle='--')
+
+
+axs[1].set_xlabel("Rang du noeud")
+axs[0].set_ylabel("Valeur du pagerank")
+axs[1].set_ylabel("Nombre de noeuds dans les sous-catégories de \"échec\" et \"box\"")
+
+
+plt.show()
-plt.show()
\ No newline at end of file