Commit: b76bd139649531e0d5bd0f6a9b1b53959639e085 - myprojects-hg-reborn (hg)

Commit MetaInfo

• Revisão: b76bd139649531e0d5bd0f6a9b1b53959639e085 (tree)
• Hora: 2011-01-25 02:01:30
• Autor: lorenzo
• Commiter: lorenzo

Mensagem de Log

Modifications to various codes: new quantities saved in the gml networks and new functions for statistics and
visualizations in the R codes.

Mudança Sumário

• modified: Python-codes/iterate_pcap2gml.py
• modified: R-codes/big_boxplot.R
• modified: R-codes/intervals_advanced.R
• modified: R-codes/intervals_reshuffled.R

Diff

diff -r 641076e85e9f -r b76bd1396495 Python-codes/iterate_pcap2gml.py
--- a/Python-codes/iterate_pcap2gml.py	Tue Jan 18 13:43:52 2011 +0000
+++ b/Python-codes/iterate_pcap2gml.py	Mon Jan 24 17:01:30 2011 +0000
@@ -9,14 +9,29 @@
 import string
 import scipy as s
 
+from sociopatterns.loader import TagMapping
+
+M = TagMapping()
+
+f = open('tag_mapping_high_presence.dat')
+for line in f:
+    (mapped_id, orig_id, t1, t2, type) = string.split(line[:-1])
+    mapped_id = int(mapped_id)
+    orig_id = int(orig_id)
+    t1 = int(t1)
+    t2 = int(t2)
+    M.add_mapping(orig_id, mapped_id, t1, t2)
+f.close()
+
+
 # ================= CONFIGURE THESE PARAMETERS ==========================
 XXTEA_CRYPTO_KEY = ( 0xf6e103d4, 0x77a739f6, 0x65eecead, 0xa40543a9 )
 DECODE = 0
 EXPERIMENT = "OBG"
-TAG_ID_MIN = 1000
-TAG_ID_MAX = 2000
-DELTAT = 20         # aggregation window
-K_MAX = 5           # filter out all nodes with degree > K_MAX (useful for filtering away clusters of tags left together)
+TAG_ID_MIN = 10
+TAG_ID_MAX = +3000
+DELTAT = 120         # aggregation window
+K_MAX = 24           # filter out all nodes with degree > K_MAX (useful for filtering away clusters of tags left together)
 # =======================================================================
 
 TAG_ID_SPAN = TAG_ID_MAX - TAG_ID_MIN
@@ -116,6 +131,9 @@
             edge_attr_list.append('%d' % (onoff[i,j] / float(cum_adj[i,j]) / 2.0) )
             edge_attr_list.append('%d' % time_first_seen[i,j])
             edge_attr_list.append('%d' % time_last_seen[i,j])
+            edge_attr_list.append('%d' % (TAG_ID_MIN+i) )
+            edge_attr_list.append('%d' % (TAG_ID_MIN+j) )
+
             edge_attr_list.append("")
 
             edge_label_list.append("edge")
@@ -130,6 +148,8 @@
             edge_label_list.append("intermittency")
             edge_label_list.append("first_seen")
             edge_label_list.append("last_seen")
+            edge_label_list.append("node_A")
+            edge_label_list.append("node_B")
             edge_label_list.append("]")
 
 
@@ -181,9 +201,15 @@
 print ("time_grid is", time_grid )
 
 loader = Loader(sys.argv[1:-1], experiment=EXPERIMENT, decode=DECODE, xxtea_crypto_key=XXTEA_CRYPTO_KEY, load_contacts=1, unique_contacts=1, load_sightings=0,
-                start_time=time_grid[0], stop_time=time_grid[-1])
+                 mapping=M,start_time=time_grid[0], stop_time=time_grid[-1])
 
-analyzer = ContactGraphAnalyzer(loader, DELTAT, get_graph=False, get_adj=True, tag_id_min=TAG_ID_MIN, tag_id_max=TAG_ID_MAX, isolated_nodes=True, weighted=False)
+# loader = Loader(sys.argv[1:], decode=0,experiment='OBG',load_contacts=1, unique_contacts=1, load_sightings=0, mapping=M)
+
+
+# analyzer = ContactGraphAnalyzer(loader, DELTAT, get_graph=False, get_adj=True, tag_id_min=TAG_ID_MIN, tag_id_max=TAG_ID_MAX, isolated_nodes=True, weighted=False)
+
+analyzer = ContactGraphAnalyzer(loader, DELTAT, get_graph=False, get_adj=True, tag_id_min=TAG_ID_MIN, tag_id_max=TAG_ID_MAX, isolated_nodes=True, weighted=True)
+
 
 
 count_day=1

diff -r 641076e85e9f -r b76bd1396495 R-codes/big_boxplot.R
--- a/R-codes/big_boxplot.R	Tue Jan 18 13:43:52 2011 +0000
+++ b/R-codes/big_boxplot.R	Mon Jan 24 17:01:30 2011 +0000
@@ -69,6 +69,8 @@
 }
 
 
+extra_norm <- c()
+
 original <- read.table("frequencies_extended.dat", header=TRUE)
 names(original) <- c("role","freq", "norm", "count_norm")
 
@@ -83,7 +85,29 @@
 
 freq_complete <- c(original$freq,extra_freq)
 
-norm_complete <- c(original$norm, extra_freq)
+
+n_ij <- read.table("n_ij.dat", header=TRUE)
+
+n_ij$role <- levels(n_ij$role)[n_ij$role]
+
+
+for (i in seq(length(extra_roles))){
+
+add_role <- extra_roles[i]
+
+sel <- which(n_ij$role==add_role)
+
+extra_norm <- c(extra_norm, n_ij$norm[sel])
+
+}
+
+
+## norm_complete <- c(original$norm, extra_freq)
+
+norm_complete <- c(original$norm, extra_norm)
+
+
+
 
 count_norm_complete <- c(original$count_norm, extra_freq)
 
@@ -1237,7 +1261,7 @@
 ## scale_y_continuous( trans="log10", limits=c(.1,3500), breaks=mb, labels=my_label2## ,expand = c(0.01,0.01)
 ##                    )+
 
-scale_y_continuous(limits=c(0,0.0025), breaks=seq(0,0.0025, by=0.0005))+
+scale_y_continuous(limits=c(0,0.026), breaks=seq(0,0.025, by=0.005))+
 
   ## scale_y_continuous( trans="log10", limits=c(.1,3500), breaks=mb, labels=my_label2)+
 
@@ -1256,6 +1280,8 @@
   opts(axis.text.x = theme_text(size=15, colour="black",vjust=1))+
   opts(axis.text.y = theme_text(size=15, colour="black", hjust=1))
 
+
+
 }
 
 
@@ -1275,6 +1301,12 @@
 
 }
 
+
+gpl_list[[1]] <- gpl_list[[1]] +
+  opts(title = expression(paste("Aggregation on 5 min, 1 hour cutoff")))+
+  opts(plot.title = theme_text(size = 20))
+
+
 fn <- paste("role-shuffle-normalized-triangles-panel.tex")
 
 width <- 15

diff -r 641076e85e9f -r b76bd1396495 R-codes/intervals_advanced.R
--- a/R-codes/intervals_advanced.R	Tue Jan 18 13:43:52 2011 +0000
+++ b/R-codes/intervals_advanced.R	Mon Jan 24 17:01:30 2011 +0000
@@ -242,6 +242,145 @@
 
 }
 
+
+
+
+
+
+
+
+t_ab_bc_improved_all_triangles <- function(data_tag,tag_id){
+
+## delta_t <- abs(diff(data_tag$ini_time))
+
+delta_t <- c()
+labels <- c()
+roles <- c()
+ternary_roles <- c()
+
+
+if (length(data_tag$id1)>1){
+
+sort_time_ini <- sort(data_tag$ini_time, decreasing=FALSE, index.return=TRUE)
+
+ms <- sort_time_ini$ix
+
+
+## sort the data once for all
+
+data_tag <- data_tag[ms, ]
+## print("tag_id is,")
+## print(tag_id)
+## print("data_tag is, ")
+## print(data_tag)
+for (i in seq(length(data_tag$id1)-1)){
+for (j in seq((i+1),length(data_tag$id1) )){
+if (((data_tag$id1[i]+data_tag$id2[i])!=(data_tag$id1[j]+data_tag$id2[j]))){
+delta_t <- c(delta_t,(data_tag$ini_time[j]-data_tag$ini_time[i]))
+labels <- rbind(labels,cbind(data_tag$id1[i],data_tag$id2[i],
+data_tag$id1[j],data_tag$id2[j]))
+
+## print("data_tag$id1[i] is, ")
+## print(data_tag$id1[i])
+
+## print("tag_id is, ")
+## print(tag_id)
+
+if (data_tag$id1[i]==tag_id){
+  tag_a <- data_tag$role_second[i]
+  tag_b <- data_tag$role_first[i]
+}else{
+    tag_a <- data_tag$role_first[i]
+    tag_b <- data_tag$role_second[i]
+}
+
+
+## print("i is, ")
+## print(i)
+## print("j is")
+## print(j)
+
+
+if (data_tag$id1[j]==tag_id){
+  tag_c <- data_tag$role_second[j]
+}else{
+    tag_c <- data_tag$role_first[j]
+}
+  
+roles <- rbind(roles,cbind(data_tag$role_first[i],data_tag$role_second[i],
+data_tag$role_first[j],data_tag$role_second[j]))
+ternary_roles <- rbind(ternary_roles,cbind(tag_a,tag_b,tag_c))
+}
+}
+}
+}
+## labels <- cbind(data_tag$id1[1:length(data_tag$id1)-1],data_tag$id2[1:length(data_tag$id2)-1],
+## data_tag$id1[2:length(data_tag$id1)],data_tag$id2[2:length(data_tag$id2)])
+return (list(delta_t, labels,roles,ternary_roles))
+}
+
+
+
+t_ab_bc_improved_all_triangles_stop <- function(data_tag,tag_id){
+
+## delta_t <- abs(diff(data_tag$ini_time))
+
+delta_t <- c()
+labels <- c()
+roles <- c()
+ternary_roles <- c()
+
+
+if (length(data_tag$id1)>1){
+
+sort_time_ini <- sort(data_tag$ini_time, decreasing=FALSE, index.return=TRUE)
+
+ms <- sort_time_ini$ix
+
+data_tag <- data_tag[ms, ]
+for (i in seq(length(data_tag$id1)-1)){
+
+
+j <- i+1
+  
+while ( (j<=length(data_tag$id1)) & ((data_tag$id1[i]+data_tag$id2[i])!=(data_tag$id1[j]+data_tag$id2[j]))){
+
+    delta_t <- c(delta_t,(data_tag$ini_time[j]-data_tag$ini_time[i]))
+labels <- rbind(labels,cbind(data_tag$id1[i],data_tag$id2[i],
+data_tag$id1[j],data_tag$id2[j]))
+
+
+if (data_tag$id1[i]==tag_id){
+  tag_a <- data_tag$role_second[i]
+  tag_b <- data_tag$role_first[i]
+}else{
+    tag_a <- data_tag$role_first[i]
+    tag_b <- data_tag$role_second[i]
+}
+
+if (data_tag$id1[j]==tag_id){
+  tag_c <- data_tag$role_second[j]
+}else{
+    tag_c <- data_tag$role_first[j]
+}
+  
+roles <- rbind(roles,cbind(data_tag$role_first[i],data_tag$role_second[i],
+data_tag$role_first[j],data_tag$role_second[j]))
+ternary_roles <- rbind(ternary_roles,cbind(tag_a,tag_b,tag_c))
+
+j <- j+1    
+    
+  }
+
+}
+}
+
+return (list(delta_t, labels,roles,ternary_roles))
+}
+
+
+
+
 get_frequence <- function(contact){
 
 freq <- table(contact)
@@ -284,11 +423,11 @@
 #######################################################################
 
 
-timeslice_dur <- 20
+timeslice_dur <- 300
 
-cut_off <- 900 #max interval durations
+cut_off <- 3600*24 #max interval durations
 
-time_reversal <- 0
+time_reversal <- 1
 
 time_uniform <- 0
 
@@ -449,9 +588,14 @@
 
 data_tag <- data_tot[sel, ]
 
-## res_single_tag <- t_ab_bc(data_tag)
 
-res_single_tag <- t_ab_bc_improved(data_tag,tag_id)
+
+## res_single_tag <- t_ab_bc_improved(data_tag,tag_id)
+
+## res_single_tag <- t_ab_bc_improved_all_triangles(data_tag,tag_id)
+
+res_single_tag <- t_ab_bc_improved_all_triangles_stop(data_tag,tag_id)
+
 
 delta_t_overall <- c(delta_t_overall, res_single_tag[[1]])
 label_overall <- rbind(label_overall,res_single_tag[[2]])

diff -r 641076e85e9f -r b76bd1396495 R-codes/intervals_reshuffled.R
--- a/R-codes/intervals_reshuffled.R	Tue Jan 18 13:43:52 2011 +0000
+++ b/R-codes/intervals_reshuffled.R	Mon Jan 24 17:01:30 2011 +0000
@@ -225,6 +225,70 @@
 }
 
 
+
+t_ab_bc_improved_all_triangles_stop <- function(data_tag,tag_id){
+
+## delta_t <- abs(diff(data_tag$ini_time))
+
+delta_t <- c()
+labels <- c()
+roles <- c()
+ternary_roles <- c()
+
+
+if (length(data_tag$id1)>1){
+
+sort_time_ini <- sort(data_tag$ini_time, decreasing=FALSE, index.return=TRUE)
+
+ms <- sort_time_ini$ix
+
+data_tag <- data_tag[ms, ]
+for (i in seq(length(data_tag$id1)-1)){
+
+
+j <- i+1
+  
+while ( (j<=length(data_tag$id1)) & ((data_tag$id1[i]+data_tag$id2[i])!=(data_tag$id1[j]+data_tag$id2[j]))){
+
+    delta_t <- c(delta_t,(data_tag$ini_time[j]-data_tag$ini_time[i]))
+labels <- rbind(labels,cbind(data_tag$id1[i],data_tag$id2[i],
+data_tag$id1[j],data_tag$id2[j]))
+
+
+if (data_tag$id1[i]==tag_id){
+  tag_a <- data_tag$role_second[i]
+  tag_b <- data_tag$role_first[i]
+}else{
+    tag_a <- data_tag$role_first[i]
+    tag_b <- data_tag$role_second[i]
+}
+
+if (data_tag$id1[j]==tag_id){
+  tag_c <- data_tag$role_second[j]
+}else{
+    tag_c <- data_tag$role_first[j]
+}
+  
+roles <- rbind(roles,cbind(data_tag$role_first[i],data_tag$role_second[i],
+data_tag$role_first[j],data_tag$role_second[j]))
+ternary_roles <- rbind(ternary_roles,cbind(tag_a,tag_b,tag_c))
+
+j <- j+1    
+    
+  }
+
+}
+}
+
+return (list(delta_t, labels,roles,ternary_roles))
+}
+
+
+
+
+
+
+
 reassign_roles <- function(data_tot, tag_role_association,tag_ids_uni, select_ids, select_roles ){
 
   reshuffled_unique_roles <- sample(tag_role_association)
@@ -258,13 +322,13 @@
 
 
 
-cut_off <- 300
+cut_off <- 3600*24
 
 
 n_iter <- 200
 
 
-sample_on_times <- 1
+sample_on_times <- 0
 
 
 data_tot <- read.table("global_data.dat", header=TRUE)
@@ -389,7 +453,9 @@
 
 ## res_single_tag <- t_ab_bc(data_tag)
 
-res_single_tag <- t_ab_bc_improved(data_tag,tag_id)
+## res_single_tag <- t_ab_bc_improved(data_tag,tag_id)
+
+res_single_tag <- t_ab_bc_improved_all_triangles_stop(data_tag,tag_id)
 
 delta_t_overall <- c(delta_t_overall, res_single_tag[[1]])
 label_overall <- rbind(label_overall,res_single_tag[[2]])