From ddb6b7b218968d11198bf0806fcb1b9a5ff6bed8 Mon Sep 17 00:00:00 2001
From: Orpaz Sondhelm <orpaz234@gmail.com>
Date: Wed, 10 Nov 2021 04:16:54 -0800
Subject: [PATCH] start

---
 .idea/vcs.xml              |   6 ++
 AnomalyDetector.h          |  27 +++++++++
 MainTrain.cpp              | 111 +++++++++++++++++++++++++++++++++++++
 SimpleAnomalyDetector.cpp  |  21 +++++++
 SimpleAnomalyDetector.h    |  38 +++++++++++++
 anomaly_detection_util.cpp |  48 ++++++++++++++++
 anomaly_detection_util.h   |  43 ++++++++++++++
 main.cpp                   |   6 --
 timeseries.cpp             |   3 +
 timeseries.h               |  47 ++++++++++++++++
 10 files changed, 344 insertions(+), 6 deletions(-)
 create mode 100644 .idea/vcs.xml
 create mode 100644 AnomalyDetector.h
 create mode 100644 MainTrain.cpp
 create mode 100644 SimpleAnomalyDetector.cpp
 create mode 100644 SimpleAnomalyDetector.h
 create mode 100644 anomaly_detection_util.cpp
 create mode 100644 anomaly_detection_util.h
 delete mode 100644 main.cpp
 create mode 100644 timeseries.cpp
 create mode 100644 timeseries.h

diff --git a/.idea/vcs.xml b/.idea/vcs.xml
new file mode 100644
index 0000000..94a25f7
--- /dev/null
+++ b/.idea/vcs.xml
@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="VcsDirectoryMappings">
+    <mapping directory="$PROJECT_DIR$" vcs="Git" />
+  </component>
+</project>
\ No newline at end of file
diff --git a/AnomalyDetector.h b/AnomalyDetector.h
new file mode 100644
index 0000000..c237f2b
--- /dev/null
+++ b/AnomalyDetector.h
@@ -0,0 +1,27 @@
+
+
+#ifndef ANOMALYDETECTOR_H_
+#define ANOMALYDETECTOR_H_
+
+#include <string>
+#include <vector>
+#include "timeseries.h"
+#include "math.h"
+using namespace std;
+
+
+class AnomalyReport{
+public:
+	const string description;
+	const long timeStep;
+	AnomalyReport(string description, long timeStep):description(description),timeStep(timeStep){}
+};
+
+class TimeSeriesAnomalyDetector {
+public:
+	virtual void learnNormal(const TimeSeries& ts)=0;
+	virtual vector<AnomalyReport> detect(const TimeSeries& ts)=0;
+	virtual ~TimeSeriesAnomalyDetector(){}
+};
+
+#endif /* ANOMALYDETECTOR_H_ */
diff --git a/MainTrain.cpp b/MainTrain.cpp
new file mode 100644
index 0000000..1572996
--- /dev/null
+++ b/MainTrain.cpp
@@ -0,0 +1,111 @@
+
+
+#include <iostream>
+#include <vector>
+#include "AnomalyDetector.h"
+#include "SimpleAnomalyDetector.h"
+#include <fstream>
+#include <stdlib.h>     /* srand, rand */
+#include <time.h>       /* time */
+#include <math.h>
+
+using namespace std;
+
+// this is a simple test to put you on the right track
+void generateTrainCSV(float a1,float b1, float a2, float b2){
+	ofstream out("trainFile1.csv");
+	out<<"A,B,C,D"<<endl;
+	Line ac(a1,b1);
+	Line bd(a2,b2);
+	for(int i=1;i<=100;i++){
+		float a=i;
+		float b=rand()%40;
+		out<<a<<","<<b<<","<<ac.f(a)-0.02+(rand()%40)/100.0f<<","<<bd.f(b)-0.02+(rand()%40)/100.0f<<endl;
+	}
+	out.close();
+}
+
+void generateTestCSV(float a1,float b1, float a2, float b2, int anomaly){
+	ofstream out("testFile1.csv");
+	out<<"A,B,C,D"<<endl;
+	Line ac(a1,b1);
+	Line bd(a2,b2);
+	for(int i=1;i<=100;i++){
+		float a=i;
+		float b=rand()%40;
+		if(i!=anomaly)
+			out<<a<<","<<b<<","<<ac.f(a)-0.02+(rand()%40)/100.0f<<","<<bd.f(b)-0.02+(rand()%40)/100.0f<<endl;
+		else
+			out<<a<<","<<b<<","<<ac.f(a)+1<<","<<bd.f(b)-0.02+(rand()%40)/100.0f<<endl;
+	}
+	out.close();
+}
+
+void checkCorrelationTrain(correlatedFeatures c,string f1, string f2, float a, float b){
+	if(c.feature1==f1){
+		if(c.feature2!=f2)
+			cout<<"wrong correlated feature of "<<f1<<" (-20)"<<endl;
+		else{
+			if(c.corrlation<0.99)
+				cout<<f1<<"-"<<f2<<" wrong correlation detected (-5)"<<endl;
+			if(c.lin_reg.a<a-0.5f || c.lin_reg.a>a+0.5f)
+				cout<<f1<<"-"<<f2<<" wrong value of line_reg.a (-5)"<<endl;
+			if(c.lin_reg.b<b-0.5f || c.lin_reg.b>b+0.5f)
+				cout<<f1<<"-"<<f2<<" wrong value of line_reg.b (-5)"<<endl;
+			if(c.threshold>0.3)
+				cout<<f1<<"-"<<f2<<" wrong threshold detected (-5)"<<endl;
+		}
+	}
+
+}
+
+int main(){
+	srand (time(NULL));
+	float a1=1+rand()%10, b1=-50+rand()%100;
+	float a2=1+rand()%20 , b2=-50+rand()%100;
+
+
+	// test the learned model: (40 points)
+	// expected correlations:
+	//	A-C: y=a1*x+b1
+	//	B-D: y=a2*x+b2
+
+	generateTrainCSV(a1,b1,a2,b2);
+	TimeSeries ts("trainFile1.csv");
+	SimpleAnomalyDetector ad;
+	ad.learnNormal(ts);
+	vector<correlatedFeatures> cf=ad.getNormalModel();
+
+	if(cf.size()!=2)
+		cout<<"wrong size of correlated features (-40)"<<endl;
+	else
+	for_each(cf.begin(),cf.end(),[&a1,&b1,&a2,&b2](correlatedFeatures c){
+		checkCorrelationTrain(c,"A","C",a1,b1); // 20 points
+		checkCorrelationTrain(c,"B","D",a2,b2); // 20 points
+	});
+
+	// test the anomaly detector: (60 points)
+	// one simply anomaly is injected to the data
+	int anomaly=5+rand()%90; // one anomaly injected in a random time step
+	generateTestCSV(a1,b1,a2,b2,anomaly);
+	TimeSeries ts2("testFile1.csv");
+	vector<AnomalyReport> r = ad.detect(ts2);
+
+	bool anomlyDetected=false;
+	int falseAlarms=0;
+	for_each(r.begin(),r.end(),[&anomaly,&anomlyDetected,&falseAlarms](AnomalyReport ar){
+		if(ar.description=="A-C" && ar.timeStep == anomaly)
+			anomlyDetected=true;
+		else
+			falseAlarms++;
+	});
+
+	if(!anomlyDetected)
+		cout<<"the anomaly was not detected (-30)"<<endl;
+
+	if(falseAlarms>0)
+		cout<<"you have "<<falseAlarms<<" false alarms (-"<<min(30,falseAlarms*3)<<")"<<endl;
+
+	cout<<"done"<<endl;
+	return 0;
+}
diff --git a/SimpleAnomalyDetector.cpp b/SimpleAnomalyDetector.cpp
new file mode 100644
index 0000000..e69c5a7
--- /dev/null
+++ b/SimpleAnomalyDetector.cpp
@@ -0,0 +1,21 @@
+
+#include "SimpleAnomalyDetector.h"
+
+SimpleAnomalyDetector::SimpleAnomalyDetector() {
+	// TODO Auto-generated constructor stub
+
+}
+
+SimpleAnomalyDetector::~SimpleAnomalyDetector() {
+	// TODO Auto-generated destructor stub
+}
+
+
+void SimpleAnomalyDetector::learnNormal(const TimeSeries& ts){
+	// TODO Auto-generated destructor stub
+}
+
+vector<AnomalyReport> SimpleAnomalyDetector::detect(const TimeSeries& ts){
+	// TODO Auto-generated destructor stub
+}
+
diff --git a/SimpleAnomalyDetector.h b/SimpleAnomalyDetector.h
new file mode 100644
index 0000000..ba5879f
--- /dev/null
+++ b/SimpleAnomalyDetector.h
@@ -0,0 +1,38 @@
+
+
+#ifndef SIMPLEANOMALYDETECTOR_H_
+#define SIMPLEANOMALYDETECTOR_H_
+
+#include "anomaly_detection_util.h"
+#include "AnomalyDetector.h"
+#include <vector>
+#include <algorithm>
+#include <string.h>
+#include <math.h>
+
+struct correlatedFeatures{
+	string feature1,feature2;  // names of the correlated features
+	float corrlation;
+	Line lin_reg;
+	float threshold;
+};
+
+
+class SimpleAnomalyDetector:public TimeSeriesAnomalyDetector{
+	vector<correlatedFeatures> cf;
+public:
+	SimpleAnomalyDetector();
+	virtual ~SimpleAnomalyDetector();
+
+	virtual void learnNormal(const TimeSeries& ts);
+	virtual vector<AnomalyReport> detect(const TimeSeries& ts);
+
+	vector<correlatedFeatures> getNormalModel(){
+		return cf;
+	}
+
+};
+
+
+
+#endif /* SIMPLEANOMALYDETECTOR_H_ */
diff --git a/anomaly_detection_util.cpp b/anomaly_detection_util.cpp
new file mode 100644
index 0000000..ebe8c94
--- /dev/null
+++ b/anomaly_detection_util.cpp
@@ -0,0 +1,48 @@
+/*
+ * animaly_detection_util.cpp
+ *
+ * Author: write your ID and name here
+ */
+
+#include <math.h>
+#include "anomaly_detection_util.h"
+
+float avg(float* x, int size){
+	return 0;
+}
+
+// returns the variance of X and Y
+float var(float* x, int size){
+	return 0;
+}
+
+// returns the covariance of X and Y
+float cov(float* x, float* y, int size){
+	return 0;
+}
+
+
+// returns the Pearson correlation coefficient of X and Y
+float pearson(float* x, float* y, int size){
+	return 0;
+}
+
+// performs a linear regression and returns the line equation
+Line linear_reg(Point** points, int size){
+
+	return Line(0,0);
+}
+
+// returns the deviation between point p and the line equation of the points
+float dev(Point p,Point** points, int size){
+	return 0;
+}
+
+// returns the deviation between point p and the line
+float dev(Point p,Line l){
+	return 0;
+}
+
+
+
+
diff --git a/anomaly_detection_util.h b/anomaly_detection_util.h
new file mode 100644
index 0000000..aabe991
--- /dev/null
+++ b/anomaly_detection_util.h
@@ -0,0 +1,43 @@
+
+#ifndef ANOMALYDETECTORUTIL_H_
+#define ANOMALYDETECTORUTIL_H_
+
+
+float avg(float* x, int size);
+
+// returns the variance of X and Y
+float var(float* x, int size);
+
+// returns the covariance of X and Y
+float cov(float* x,  float* y, int size);
+
+
+// returns the Pearson correlation coefficient of X and Y
+float pearson(float* x, float* y, int size);
+
+class Line{
+public:
+	float a,b;
+	Line():a(0),b(0){};
+	Line(float a, float b):a(a),b(b){}
+	float f(float x){
+		return a*x+b;
+	}
+};
+
+class Point{
+public:
+	float x,y;
+	Point(float x, float y):x(x),y(y){}
+};
+
+// performs a linear regression and returns the line equation
+Line linear_reg(Point** points, int size);
+
+// returns the deviation between point p and the line equation of the points
+float dev(Point p,Point** points, int size);
+
+// returns the deviation between point p and the line
+float dev(Point p,Line l);
+
+#endif
diff --git a/main.cpp b/main.cpp
deleted file mode 100644
index bc8f460..0000000
--- a/main.cpp
+++ /dev/null
@@ -1,6 +0,0 @@
-#include <iostream>
-
-int main() {
-    std::cout << "Hello, World!" << std::endl;
-    return 0;
-}
diff --git a/timeseries.cpp b/timeseries.cpp
new file mode 100644
index 0000000..59275f2
--- /dev/null
+++ b/timeseries.cpp
@@ -0,0 +1,3 @@
+#include "timeseries.h"
+
+
diff --git a/timeseries.h b/timeseries.h
new file mode 100644
index 0000000..a70e84f
--- /dev/null
+++ b/timeseries.h
@@ -0,0 +1,47 @@
+#include <fstream>
+#include <vector>
+//#include <string>
+
+#ifndef TIMESERIES_H_
+#define TIMESERIES_H_
+
+using namespace std;
+
+class TimeSeries{
+
+public:
+
+    TimeSeries(const char* CSVfileName){
+        string line;
+        ifstream myFile (CSVfileName);
+        vector<(pair<(string), (vector<float)>)> columns;
+        int flag = 1;
+        if (myFile.is_open())
+        {
+            while ( getline (myFile,line) )
+            {
+                std::istringstream ss(line);
+                std::string token
+                while(std::getline(ss, token, ',')) {
+                    if (flag == 1) {
+                        vector<float> v1
+                        columns.push_bac(make_pair(token, v1));
+
+                    }
+                    else{
+                        for(int i = 0, i < columns.size(), i++){
+                            columns[i].second = token;
+                        }
+                    }
+                }
+                flag = 0;
+
+            }
+            myFile.close();
+        }
+    }
+};
+
+
+
+#endif /* TIMESERIES_H_ */