From 328453e61a8992729c0ecc661ad324b63af6b41d Mon Sep 17 00:00:00 2001
From: Tim Redfern <tim@eclectronics.org>
Date: Wed, 11 Dec 2013 13:49:45 +0000
Subject: tweaking intensity segmenter

---
 rotord/src/nodes_audio_analysis.cpp | 137 +++++++++++++++++++++++++-----------
 1 file changed, 96 insertions(+), 41 deletions(-)

(limited to 'rotord/src')

diff --git a/rotord/src/nodes_audio_analysis.cpp b/rotord/src/nodes_audio_analysis.cpp
index 1bc82e6..96244d0 100644
--- a/rotord/src/nodes_audio_analysis.cpp
+++ b/rotord/src/nodes_audio_analysis.cpp
@@ -113,6 +113,9 @@ namespace Rotor{
 	bool sortseggrps(std::pair<double,vector<pair<double,int> > > i,std::pair<double,vector<pair<double,int> > > j){
 		return (i.first<j.first);
 	}
+	bool sortgroupmembers(pair<double,int> i,pair<double,int> j){
+		return (i.first<j.first);
+	}
 	void Intensity_segmenter::cleanup(){
 		//algorithm idea:
 		//get average tempo and intensity for each segment and store them
@@ -157,7 +160,14 @@ namespace Rotor{
 			}
 			i++;
 		}
-		for (auto s:similarities) cerr<<"group "<<s.first<<" count: "<<s.second.size()<<endl;
+		for (auto s:similarities) {
+			string list="";
+			for (int j=0;j<s.second.size();j++){
+				if (j>0) list+=",";
+				list +=toString(s.second[j]);
+			}
+			cerr<<"group "<<s.first<<" ["<<list<<"]"<<endl;
+		}
 
 		cerr<<analysers["segmenter"].features.size()<<" segments"<<endl;
 		cerr<<analysers["tempo"].features.size()<<" tempo features"<<endl;
@@ -172,7 +182,6 @@ namespace Rotor{
 		vector<double> times;
 		auto g=++analysers["segmenter"].features.begin();
 		for (auto f=analysers["segmenter"].features.begin();g!=analysers["segmenter"].features.end();f++,g++,i++){
-			cerr<<"segment "<<i<<": "<<f->first<<" to "<<g->first<<endl;
 			times.push_back(f->first);
 			//integrate tempo and intensity algorithmically
 			double tempo=0;
@@ -190,8 +199,7 @@ namespace Rotor{
 			if (tempo>max_tempo) max_tempo=tempo;
 			if (tempo<min_tempo) min_tempo=tempo;
 			tempos.push_back(tempo);
-			cerr<<"segment "<<i<<" average tempo: "<<tempo<<endl;
-
+			
 			double intensity=0;
 			if (analysers["intensity"].features.size()) {
 				double pt=f->first;
@@ -207,7 +215,21 @@ namespace Rotor{
 			if (intensity>max_intensity) max_intensity=intensity;
 			if (intensity<min_intensity) min_intensity=intensity;
 			intensities.push_back(intensity);
+
+			cerr<<"segment "<<i<<": "<<f->first<<" to "<<g->first<<" average tempo: "<<tempo<<" ,intensity: "<<intensity<<" ,weighted: "<<(tempo*parameters["tempo_weight"]->value)+(intensity*parameters["intensity_weight"]->value)<<endl;
 		}
+		//
+		//
+		//need to calculate the last segment
+		//
+		//
+		//either by adding a bit of code here or by adding a dummy feature at the track duration, previously
+		//
+		//
+		//
+		//
+
+
 		//make relative scale 0.0-1.0 and save weighted totals
 		vector< pair<int,double>> totals;
 		vector<double> totalsmap;
@@ -298,56 +320,89 @@ nned to retrieve total intensity by segment
 		std::sort(seggrps.begin(),seggrps.end(),sortseggrps);
 		//possible mergers will be with groups with adjacent intensity
 		if (((int)parameters["levels"]->value)>0) {
-			while (seggrps.size()>(int)parameters["levels"]->value){
-				//reduce similarity groups
-				//decide the best 2 to merge
-				vector<double> diffs;
-				for (int j=0;j<seggrps.size()-1;j++) diffs.push_back(seggrps[j+1].first-seggrps[j].first);
-				int smallest=0;
-				for (int j=1;j<diffs.size();j++) if (diffs[j]<diffs[smallest]) smallest=j;
-				for (int j=0;j<seggrps[smallest].second.size();j++) {
-					seggrps[smallest+1].second.push_back(seggrps[smallest].second[j]);
-				}
-				//recalculate intensity average
-				double avg=0.0f;
-				for (auto p:seggrps[smallest+1].second) avg+=p.first;
-				avg/=seggrps[smallest+1].second.size();
-				seggrps[smallest+1].first=avg;
+			if (seggrps.size()>(int)parameters["levels"]->value){
+				while (seggrps.size()>(int)parameters["levels"]->value){
+					//reduce similarity groups
+					//decide the best 2 to merge
+					vector<double> diffs;
+					for (int j=0;j<seggrps.size()-1;j++) diffs.push_back(seggrps[j+1].first-seggrps[j].first);
+					int smallest=0;
+					for (int j=1;j<diffs.size();j++) if (diffs[j]<diffs[smallest]) smallest=j;
+					for (int j=0;j<seggrps[smallest].second.size();j++) {
+						seggrps[smallest+1].second.push_back(seggrps[smallest].second[j]);
+					}
+					//recalculate intensity average
+					double avg=0.0f;
+					for (auto p:seggrps[smallest+1].second) avg+=p.first;
+					avg/=seggrps[smallest+1].second.size();
+					seggrps[smallest+1].first=avg;
 
-				seggrps.erase(seggrps.begin()+smallest);
+					seggrps.erase(seggrps.begin()+smallest);
+				}
+				cerr<<"intensities merged, "<<seggrps.size()<<" levels remain"<<endl;
 			}
-			cerr<<"intensities merged, "<<seggrps.size()<<" levels remain"<<endl;
-			while (false) {
-			//while (seggrps.size()<min((int)parameters["levels"]->value,(int)totalsmap.size())) {
-				//split groups
-				//calculate standard deviation of intensity variation
-				vector<double> devs;
-				for (int j=0;j<seggrps.size()-1;j++) {
-					double avg=0.0;
-					double dev=0.0;
-					for (int k=0;k<seggrps[j].second.size();k++){
-						avg+=k<seggrps[j].second[k].first;
+			if  (seggrps.size()<min((int)parameters["levels"]->value,(int)totalsmap.size())){
+				while (seggrps.size()<min((int)parameters["levels"]->value,(int)totalsmap.size())) {
+					//split groups
+					//calculate standard deviation of intensity variation
+					vector<double> devs;
+					for (int j=0;j<seggrps.size()-1;j++) {
+						double avg=0.0;
+						double dev=0.0;
+						for (int k=0;k<seggrps[j].second.size();k++){
+							avg+=k<seggrps[j].second[k].first;
+						}
+						avg/=seggrps[j].second.size();
+						for (int k=0;k<seggrps[j].second.size();k++){
+							dev+=pow(avg-k<seggrps[j].second[k].first,2.0);
+						}
+						dev/=seggrps[j].second.size();
+						devs.push_back(pow(dev,0.5));
 					}
-					avg/=seggrps[j].second.size();
-					for (int k=0;k<seggrps[j].second.size();k++){
-						dev+=pow(avg-k<seggrps[j].second[k].first,2.0);
+					//find group with largest standard deviation
+					int largest=0;
+					for (int j=1;j<devs.size();j++) if (devs[j]>devs[largest]) largest=j;
+					//sanity check: if there are any groups that can be split they will have larger SD than singleton groups
+					//sort members of the group
+					std::sort(seggrps[largest].second.begin(),seggrps[largest].second.end(),sortgroupmembers);
+					//create a new group
+					std::pair<double,vector<pair<double,int> > > newgroup;
+					for (int j=seggrps[largest].second.size();j>seggrps[largest].second.size()/2;j--) {
+						newgroup.second.push_back(seggrps[largest].second[j]);
+						seggrps[largest].second.erase(seggrps[largest].second.begin()+j);
 					}
-					dev/=seggrps[j].second.size();
-					devs.push_back(pow(dev,0.5));
+
+					//refresh averages for the 2 groups
+					double avg=0.0f;
+					for (auto p:seggrps[largest].second) avg+=p.first;
+					avg/=seggrps[largest].second.size();
+					seggrps[largest].first=avg;
+
+					avg=0.0f;
+					for (auto p:newgroup.second) avg+=p.first;
+					avg/=newgroup.second.size();
+					newgroup.first=avg;
+
+					//add the new group
+					seggrps.push_back(newgroup);
 				}
-				//find group with largest standard deviation
-				int largest=0;
-				for (int j=1;j<devs.size();j++) if (devs[j]>devs[largest]) largest=j;
-				//sanity check: if there are any groups that can be split they will have larger SD than singleton groups
-				//TODO actually split group
+				
+
 			}
+			cerr<<"similaritity groups split, "<<seggrps.size()<<" levels total"<<endl;
+			//seggrps are now out of order
+			std::sort(seggrps.begin(),seggrps.end(),sortseggrps);
 		}
 
 		map<int,int> outputvalues;
 		for (int j=0;j<seggrps.size();j++){
+			string list="";
 			for (int k=0;k<seggrps[j].second.size();k++){
 				outputvalues[seggrps[j].second[k].second]=j;
+				if (k>0) list+=",";
+				list +=toString(seggrps[j].second[k].second);
 			}
+			cerr<<"output value: "<<j<<" assigned to ["<<list<<"]"<<endl;
 		}
 
 
-- 
cgit v1.2.3