Browse Source

Added corpus description table

William N. Havard 1 year ago
parent
commit
04d7c7f6be
3 changed files with 31 additions and 0 deletions
  1. 1 0
      .gitignore
  2. 23 0
      CODE/all-analyses.Rmd
  3. 7 0
      OUTPUT/corpus_description.csv

+ 1 - 0
.gitignore

@@ -4,3 +4,4 @@ all-analyses.log
 .Rhistory
 .idea
 .RData
+.~*

+ 23 - 0
CODE/all-analyses.Rmd

@@ -208,7 +208,30 @@ for (data_set in data_sets){   # data_set = "aclew"
 
 
 write.csv(df.icc.corpus,"../OUTPUT/df.icc.corpus.csv",row.names=F)
+```
 
+```{r}
+mydat <- read.csv(paste0('../DATA/', 'aclew','_metrics.csv'))
+
+child_per_corpus = setNames(aggregate(data = mydat, child_id ~ experiment, function(child_id) length(unique(child_id))), c('experiment', 'No_Children'))
+rec_per_corpus = setNames(aggregate(data = mydat, session_id ~ experiment, function(session_id) length(unique(session_id))), c('experiment', 'No_Rec'))
+dur_per_corpus = setNames(aggregate(data = mydat, duration_vtc ~ experiment, function(duration_vtc) sum(duration_vtc)/3.6e+6), c('experiment', 'Duration_h'))
+age_mean_per_corpus = setNames(aggregate(data = mydat, age ~ experiment, function(age) mean(age)), c('experiment', 'Mean_Age'))
+age_min_per_corpus = setNames(aggregate(data = mydat, age ~ experiment, function(age) min(age)), c('experiment', 'Min_Age'))
+age_max_per_corpus = setNames(aggregate(data = mydat, age ~ experiment, function(age) max(age)), c('experiment', 'Max_Age'))
+
+corp_code = data.frame(
+  experiment=c("bergelson", "cougar", "fausey-trio", "lucid", "warlaumont", "winnipeg"),
+  code=c("BER", "COU", "TRI", "L05", "WAR", "MCD"),
+  location=c("Northeast US", "Northwest US", "Western US", "Northwest England", "Western US", "Western Canada")
+)
+
+corp_desc_list = list(corp_code, child_per_corpus, rec_per_corpus, dur_per_corpus, age_mean_per_corpus, age_min_per_corpus, age_max_per_corpus)
+
+corpus_description <- Reduce(function(x, y) merge(x, y, all=TRUE), corp_desc_list)
+corpus_description <- transform(corpus_description, Age_Range=paste(Min_Age, Max_Age, sep="-"))
+corpus_description <- subset(corpus_description, select = -c(Min_Age, Max_Age))
+write.csv(corpus_description, "../OUTPUT/corpus_description.csv", sep='\t')
 nkids=length(levels(factor(paste(mydat$experiment,mydat$child_id))))
 ```
 

+ 7 - 0
OUTPUT/corpus_description.csv

@@ -0,0 +1,7 @@
+"","experiment","code","location","No_Children","No_Rec","Duration_h","Mean_Age","Age_Range"
+"1","bergelson","BER","Northeast US",44,522,7319.367525,11.227969348659,"6-17"
+"2","cougar","COU","Northwest US",34,255,2762.153115,26.6862745098039,"0-66"
+"3","fausey-trio","TRI","Western US",28,84,1147.4797825,8.94047619047619,"6-12"
+"4","lucid","L05","Northwest England",35,235,3624.28059166667,20.0340425531915,"10-31"
+"5","warlaumont","WAR","Western US",15,40,498.573130555556,6.15,"2-18"
+"6","winnipeg","MCD","Western Canada",13,44,370.626272222222,13.5227272727273,"2-33"