diff --git a/.RData b/.RData
index 5567909..fd1d1c7 100644
Binary files a/.RData and b/.RData differ
diff --git a/.Rhistory b/.Rhistory
index ea91c6f..fa112b9 100644
--- a/.Rhistory
+++ b/.Rhistory
@@ -1,512 +1,512 @@
-dm <- dim(GAX)
-dm
-colnames(GAX)[dm[2]]
-colnames(GAX)[dm[2]] <- "Direction"
-colnames(GAX)
-issd <- "2013-01-01"
-ised <- "2017-12-31"
-ossd <- "2018-01-01"
-osed <- "2018-11-20"
-isrow <- which(index(GAX) >= issd & index(GAX) <= ised)
-osrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-isGAX <- GAX[isrow,]
-osGAX <- GAX[osrow,]
-#표준화#
-isme <- apply(isGAX, 2, mean)
-isstd <- apply(isGAX, 2, sd)
-isidn <- matrix(1, dim(isGAX)[1], dim(isGAX)[2])
-norm_isGAX <- (isGAX - t(isme * t(isGAX))) / t(isstd * t(isidn))
-dm<-dim(isGAX)
-norm_isGAX[,dm[2]] <- direction[isrow]
-formula <- paste("Direction ~ .", sep ="")
-model <-glm(formula, family = "binomial", norm_isGAX)
-summary(model)
-pred <- predict(model, norm_isGAX)
-prob <- 1/(1+exp(-(pred)))
-par(mflow = c(2,1))
-plot(pred,type = "l")
-plot(prob, type = "l")
-head(prob)
-tail(prob)
-pred_direction <- NULL
-pred_direction[prob > 0.5] <- 1
-pred_direction[prob <= 0.5] <- 0
-pred_direction
-#오차행렬에 집어넣을 때, pred_direction이랑 norm_isdji 간 객체 차 존재. Table로 집어넣어야 함, e1071패키지 설치 후 실행
-matrix <- confusionMatrix(table(pred_direction,norm_isGAX$Direction))
-matrix
-#예측 정확도 94%
-#정규화
-osidn <- matrix(1,dim(osGAX)[1], dim(osGAX)[2])
-norm_osGAX <- (osGAX - t(isme*t(osidn))) / t(isstd*t(osidn))
-dm <- dim(osGAX)
-norm_osGAX[,dm[2]] <- direction[osrow]
-#표본 외 데이터 값, 확률
-ospred <- predict(model, norm_osGAX)
-osprob <- 1/(1+exp(-(ospred)))
-ospred_direction <- NULL
-ospred_direction[osprob >0.5]<-1
-ospred_direction[osprob<=0.5]<-0
-osmatrix <- confusionMatrix(table(ospred_direction, norm_osGAX$Direction))
-osmatrix
-#84% accuracy, 트레이딩 비용, 시장 슬리피지 고려X, 다른 전략 X 오로지 예측
-#### 트레이딩 후 Cummulative return 구하는 코딩
-#로지스틱 CUM RET 구하는 법
-signal <- ifelse(ospred_direction == 1, 1, ifelse(ospred_direction == 0,-1,0))
-testrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-GAX <- GOLD.AX$GOLD.AX.Close
-ret <- GAX/lag(GAX) -1
-ret <- ret[testrow]
-ret
-cost <- 0
-length(signal)
-length(ret)
-dim(ret)
-length(testrow)
-trade_ret <- ret * Lag(signal) - cost
-dim(ret)
-cumm_ret <- Return.cumulative(trade_ret)
-anual_ret <- Return.annualized(trade_ret)
-charts.PerformanceSummary(trade_ret)
-### Data ###
-getSymbols("GOLD.AX", src = "yahoo")
-GAX <- GOLD.AX
-GAX <- GAX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-#Varaible #
-avg10 <- rollapply(GAX, 10, mean)
-avg20 <- rollapply(GAX, 20, mean)
-std10 <- rollapply(GAX,10,sd)
-std20 <- rollapply(GAX,20,sd)
-rsi5 <- RSI(GAX, 5, "SMA")
-rsi14 <- RSI(GAX, 14, "SMA")
-macd12269 <- MACD(GAX, 12, 26, 9, "SMA")
-macd7205 <- MACD(GAX, 7, 20, 5, "SMA")
-bbands <- BBands(GAX, 20, "SMA", 2)
-direction <- NULL
-direction[GAX > Lag(GAX, 20)] <- 1
-direction[GAX < Lag(GAX, 20)] <- 0
-GAX <- cbind(GAX, avg10, avg20, std10, std20, rsi5, rsi14, macd12269, macd7205, bbands, direction)
-dm <- dim(GAX)
-dm
-colnames(GAX)[dm[2]]
-colnames(GAX)[dm[2]] <- "Direction"
-colnames(GAX)
-issd <- "2013-01-01"
-ised <- "2017-12-31"
-ossd <- "2018-01-01"
-osed <- "2018-11-20"
-isrow <- which(index(GAX) >= issd & index(GAX) <= ised)
-osrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-isGAX <- GAX[isrow,]
-osGAX <- GAX[osrow,]
-#표준화#
-isme <- apply(isGAX, 2, mean)
-isstd <- apply(isGAX, 2, sd)
-isidn <- matrix(1, dim(isGAX)[1], dim(isGAX)[2])
-norm_isGAX <- (isGAX - t(isme * t(isGAX))) / t(isstd * t(isidn))
-dm<-dim(isGAX)
-norm_isGAX[,dm[2]] <- direction[isrow]
-formula <- paste("Direction ~ .", sep ="")
-model <-glm(formula, family = "binomial", norm_isGAX)
-summary(model)
-pred <- predict(model, norm_isGAX)
-prob <- 1/(1+exp(-(pred)))
-par(mflow = c(2,1))
-plot(pred,type = "l")
-plot(prob, type = "l")
-head(prob)
-tail(prob)
-pred_direction <- NULL
-pred_direction[prob > 0.5] <- 1
-pred_direction[prob <= 0.5] <- 0
-pred_direction
-#오차행렬에 집어넣을 때, pred_direction이랑 norm_isdji 간 객체 차 존재. Table로 집어넣어야 함, e1071패키지 설치 후 실행
-matrix <- confusionMatrix(table(pred_direction,norm_isGAX$Direction))
-matrix
-#예측 정확도 94%
-#정규화
-osidn <- matrix(1,dim(osGAX)[1], dim(osGAX)[2])
-norm_osGAX <- (osGAX - t(isme*t(osidn))) / t(isstd*t(osidn))
-dm <- dim(osGAX)
-norm_osGAX[,dm[2]] <- direction[osrow]
-#표본 외 데이터 값, 확률
-ospred <- predict(model, norm_osGAX)
-osprob <- 1/(1+exp(-(ospred)))
-ospred_direction <- NULL
-ospred_direction[osprob >0.5]<-1
-ospred_direction[osprob<=0.5]<-0
-osmatrix <- confusionMatrix(table(ospred_direction, norm_osGAX$Direction))
-osmatrix
-#84% accuracy, 트레이딩 비용, 시장 슬리피지 고려X, 다른 전략 X 오로지 예측
-#### 트레이딩 후 Cummulative return 구하는 코딩
-#로지스틱 CUM RET 구하는 법
-signal <- ifelse(ospred_direction == 1, 1, ifelse(ospred_direction == 0,-1,0))
-testrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-GAX <- GOLD.AX$GOLD.AX.Close
-ret <- GAX/lag(GAX) -1
-ret <- ret[testrow]
-ret
-cost <- 0
-length(signal)
-length(ret)
-dim(ret)
-length(testrow)
-trade_ret <- ret * Lag(signal) - cost
-dim(ret)
-cumm_ret <- Return.cumulative(trade_ret)
-anual_ret <- Return.annualized(trade_ret)
-charts.PerformanceSummary(trade_ret)
-cumm_ret
-anual_ret
-### Neural Network Machine Learning ###
-rm(list=ls())
-getSymbols("GOLD.AX", src = "yahoo")
-GAX <- GOLD.AX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-chartSeries(ClCl(GAX))
-plot(GAX)
-#수익률 계산, 델타 함수가 안되어요 ㅜㅜ
-ret <- GAX/lag(GAX) -1 # 중요함
-avg10 <- rollapply(GAX,10,mean)
-avg20 <- rollapply(GAX,20,mean)
-std10<-rollapply(GAX,10,sd)
-std20<-rollapply(GAX,20,sd)
-rsi5 <- RSI(GAX,5,"SMA")
-rsi14<-RSI(GAX,14,"SMA")
-macd12269 <- MACD(GAX,12,26,9,"SMA")
-macd7205 <- MACD(GAX,7,20,5,"SMA")
-bbands<-BBands(GAX,20,"SMA",2)
-#지난 20일간 수익률 2% 이상이면 up, -2% down, 그사이 nowhere
-direction <- data.frame(matrix(NA,dim(GAX)[1],1))
-#20일 수익률
-lagret <- (GAX - Lag(GAX,20)) / Lag(GAX,20)
-direction[lagret > 0.02] <- "Up"
-direction[lagret < -0.02] <- "Down"
-direction[lagret < 0.02 & lagret > -0.02] <- "Nowhere"
-GAX <- cbind(GAX,avg10,avg20, std10, std20,rsi5,rsi14,macd12269,macd7205,bbands)
-#훈련용, 검증용, 평가용 데이터집합
-train_sdate <- "2013-01-01"
-train_edate <- "2016-12-31"
-vali_sdate <- "2017-01-01"
-vali_edate <- "2017-12-31"
-test_sdate <- "2018-01-01"
-test_edate <- "2018-12-31"
-trainrow <- which(index(GAX) >= train_sdate & index(GAX) <= train_edate)
-valirow <- which(index(GAX) >= vali_sdate & index(GAX) <= vali_edate)
-testrow <- which(index(GAX) >= test_sdate & index(GAX) <= test_edate)
-trainGAX <- GAX[trainrow,]
-valiGAX <- GAX[valirow,]
-testGAX <- GAX[testrow,]
-trainme <- apply(trainGAX,2,mean)
-trainstd <- apply(trainGAX,2,sd)
-#정규화
-trainidn <- (matrix(1,dim(trainGAX)[1],dim(trainGAX)[2]))
-valiidn <- (matrix(1,dim(valiGAX)[1],dim(valiGAX)[2]))
-testidn <- (matrix(1,dim(testGAX)[1],dim(testGAX)[2]))
-norm_trainGAX <- (trainGAX - t(trainme*t(trainidn)))/t(trainstd*t(trainidn))
-norm_valiGAX <- (valiGAX - t(trainme*t(valiidn)))/t(trainstd*t(valiidn))
-norm_testGAX <- (testGAX - t(trainme*t(testidn)))/t(trainstd*t(testidn))
-traindir <- direction[trainrow,1]
-validir <- direction[valirow,1]
-testdir<-direction[testrow,1]
-#신경망 적합(정규화열,날짜별 방향, 신경수, 트레이스 출력 여부)
-set.seed(1)
-model <- nnet(norm_trainGAX, class.ind(traindir),size = 4, trace = F)
-model
-dim(norm_trainGAX)
-vali_pred <- predict(model, norm_valiGAX)
-head(vali_pred)
-vali_pred_class <- data.frame(matrix(NA,dim(vali_pred)[1],1))
-vali_pred_class[vali_pred[,"Down"]>0.5, 1]<- "Down"
-vali_pred_class[vali_pred[,"Nowhere"]>0.5,1] <- "Nowhere"
-vali_pred_class[vali_pred[,"Up"]>0.5, 1] <- "Up"
-vali_pred_class
-matrix<- confusionMatrix(table(vali_pred_class[,1],validir))
-matrix
-#87.65%
-test_pred <- predict(model,norm_testGAX)
-test_pred_class <- data.frame(matrix(NA,dim(test_pred)[1],1))
-test_pred_class[test_pred[,"Down"]>0.5, 1]<- "Down"
-test_pred_class[test_pred[,"Nowhere"]>0.5,1] <- "Nowhere"
-test_pred_class[test_pred[,"Up"]>0.5, 1] <- "Up"
-test_matrix <- confusionMatrix(table(test_pred_class[,1],testdir))
-test_matrix
-#82pro
-#Signal generator
-signal <- ifelse(test_pred_class == "Up", 1, ifelse(test_pred_class == "Down",-1,0))
-ret <- ret[testrow]
-ret
-cost <- 0
-trade_ret <- ret * Lag(signal) - cost
-cumm_ret <- Return.cumulative(trade_ret)
-anual_ret <- Return.annualized(trade_ret)
-charts.PerformanceSummary(trade_ret)
-#charts.PerformanceSummary(cumm_ret)
-#plot(cumm_ret)
-#Deep Neural Network
-set.seed(1)
-model <- dbn.dnn.train(norm_testGAX,class.ind(traindir),hidden=c(3,4,6))
-nn.predict(model,norm_valiGAX)
-nn.test(model,norm_valiGAX,class.ind(validir),t=0.4)
-data <- cbind(as.data.frame(norm_trainGAX),traindir)
-class(norm_trainGAX)
-class(traindir)
-h2o.init()
-datah2o <- as.h2o(data,"h2o")
-class(datah2o)
-dim(datah2o)
-#은닉층이 4개, 각 뉴런수 4,5,2,7
-model <- h2o.deeplearning(1:15, 16,training_frame = datah2o,hidden = c(4,5,2,7))
-vali_pred <- predict(model, as.h2o(norm_valiGAX,"h2o"))
-vali_pred <- as.data.frame(vali_pred)
-vali_pred_class <- data.frame(matrix(NA,dim(vali_pred)[1],1))
-vali_pred_class[vali_pred[,"Down"]>0.5, 1]<- "Down"
-vali_pred_class[vali_pred[,"Nowhere"]>0.5,1] <- "Nowhere"
-vali_pred_class[vali_pred[,"Up"]>0.5, 1] <- "Up"
-vali_matrix <- confusionMatrix(table(vali_pred_class[,1],validir))
-vali_matrix
-## Deep Graph ##
-signal <- ifelse(vali_pred_class == "Up", 1, ifelse(vali_pred_class == "Down",-1,0))
-ret <- GAX/lag(GAX) -1 # 중요함
-ret <- ret[valirow]
-ret <- ret[,1]
-cost <- 0
-trade_ret <- ret * Lag(signal) - cost
-cumm_ret <- Return.cumulative(trade_ret)
-anual_ret <- Return.annualized(trade_ret)
-charts.PerformanceSummary(trade_ret)
-rm(list=ls())
-getSymbols("GOLD.AX", src = "yahoo")
-GAX <- GOLD.AX
-GAX <- GAX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-avg10 <- rollapply(GAX, 10, mean)
-avg20 <- rollapply(GAX, 20, mean)
-std10 <- rollapply(GAX,10,sd)
-std20 <- rollapply(GAX,20,sd)
-rsi5 <- RSI(GAX, 5, "SMA")
-rsi14 <- RSI(GAX, 14, "SMA")
-macd12269 <- MACD(GAX, 12, 26, 9, "SMA")
-macd7205 <- MACD(GAX, 7, 20, 5, "SMA")
-bbands <- BBands(GAX, 20, "SMA", 2)
-direction <- NULL
-direction[GAX > Lag(GAX, 20)] <- 1
-direction[GAX < Lag(GAX, 20)] <- 0
-GAX <- cbind(GAX, avg10, avg20, std10, std20, rsi5, rsi14, macd12269, macd7205, bbands, direction)
-dm <- dim(GAX)
-dm
-colnames(GAX)[dm[2]]
-colnames(GAX)[dm[2]] <- "Direction"
-colnames(GAX)
-issd <- "2013-01-01"
-ised <- "2017-12-31"
-ossd <- "2018-01-01"
-osed <- "2018-12-31"
-isrow <- which(index(GAX) >= issd & index(GAX) <= ised)
-osrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-isGAX <- GAX[isrow,]
-osGAX <- GAX[osrow,]
-#표준화
-isme <- apply(isGAX, 2, mean)
-isstd <- apply(isGAX, 2, sd)
-isidn <- matrix(1, dim(isGAX)[1], dim(isGAX)[2])
-norm_isGAX <- (isGAX - t(isme * t(isidn))) / t(isstd * t(isidn))
-norm_isGAX
-dm<-dim(isGAX)
-norm_isGAX[,dm[2]] <- direction[isrow]
-#정규화
-osidn <- matrix(1,dim(osGAX)[1], dim(osGAX)[2])
-norm_osGAX <- (osGAX - t(isme*t(osidn))) / t(isstd*t(osidn))
-dm <- dim(osGAX)
-norm_osGAX[,dm[2]] <- direction[osrow]
-#방향설정
-GAX <- GOLD.AX
-GAX <- GAX[,"GOLD.AX.Close"] #다시 설정
-GAX <- na.omit(GAX)
-lagret <- (GAX - Lag(GAX,20))/Lag(GAX,20)
-direction[lagret >0.02]<-"Up"
-direction[lagret < -0.02]<-"Down"
-direction[lagret < 0.02 & lagret > -0.02]<-"Nowhere"
-isdir <- direction[isrow]
-osdir <- direction[osrow]
-model <- svm(norm_isGAX, as.factor(isdir))
-model
-pred <- predict(model, norm_osGAX)
-head(pred)
-table(pred,osdir)
-model
-#적중률
-sum(diag(table(pred,osdir)))/sum(table(pred,osdir))
-####서포트벡터 트레이딩 후 Cummulative return 구하는 코딩
-signal <- ifelse(pred == "Up", 1, ifelse(pred == "Down",-1,0))
-#테스트할 날짜 데이터 집합
-testrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-testrow
-#리턴 구하는 공식과 그 공식 활용 테스트 날짜별 리턴
-ret <- GAX/lag(GAX) -1
-ret <- ret[testrow]
-ret
-cost <- 0
-trade_ret <- ret * Lag(signal) - cost
-length(signal)
-dim(ret)
-cumm_ret <- Return.cumulative(trade_ret)
-anual_ret <- Return.annualized(trade_ret)
-charts.PerformanceSummary(trade_ret)
-### K-clustering ###
-rm(list=ls())
-#Data
-getSymbols("GOLD.AX", src = "yahoo")
-GAX <- GOLD.AX
-GAX <- GAX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-#variable
-avg10 <- rollapply(GAX, 10, mean)
-avg20 <- rollapply(GAX, 20, mean)
-std10 <- rollapply(GAX,10,sd)
-std20 <- rollapply(GAX,20,sd)
-rsi5 <- RSI(GAX, 5, "SMA")
-rsi14 <- RSI(GAX, 14, "SMA")
-macd12269 <- MACD(GAX, 12, 26, 9, "SMA")
-macd7205 <- MACD(GAX, 7, 20, 5, "SMA")
-bbands <- BBands(GAX, 20, "SMA", 2)
-direction <- NULL
-direction[GAX > Lag(GAX, 20)] <- 1
-direction[GAX < Lag(GAX, 20)] <- 0
-GAX <- cbind(GAX, avg10, avg20, std10, std20, rsi5, rsi14, macd12269, macd7205, bbands, direction)
-dm <- dim(GAX)
-dm
-colnames(GAX)[dm[2]]
-colnames(GAX)[dm[2]] <- "Direction"
-colnames(GAX)
-issd <- "2013-01-01"
-ised <- "2017-12-31"
-ossd <- "2018-01-01"
-osed <- "2018-12-31"
-isrow <- which(index(GAX) >= issd & index(GAX) <= ised)
-osrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-isGAX <- GAX[isrow,]
-osGAX <- GAX[osrow,]
-#표준화
-isme <- apply(isGAX, 2, mean)
-isstd <- apply(isGAX, 2, sd)
-isidn <- matrix(1, dim(isGAX)[1], dim(isGAX)[2])
-norm_isGAX <- (isGAX - t(isme * t(isidn))) / t(isstd * t(isidn))
-dm<-dim(isGAX)
-norm_isGAX[,dm[2]] <- direction[isrow]
-osidn <- matrix(1,dim(osGAX)[1], dim(osGAX)[2])
-norm_osGAX <- (osGAX - t(isme*t(osidn))) / t(isstd*t(osidn))
-dm <- dim(osGAX)
-norm_osGAX[,dm[2]] <- direction[osrow]
-clusters <- 3
-set.seed(1)
-#디렉션 제거
-dm<- dim(isGAX)
-isGAX[,-dm[2]]
-isGAX <- isGAX[,-dm[2]]
-norm_isGAX <- norm_isGAX[,-dm[2]]
-dm<-dim(osGAX)
-osGAX<-osGAX[,-dm[2]]
-norm_osGAX<-norm_osGAX[,-dm[2]]
-#isdji 바탕으로 클러스터화 모델
-model <- kmeans(norm_isGAX, clusters)
-head(model$cluster)
-model$cluster
-model$center
-model$size
-#클러스터 내제곱합과 총제곱합 비율 최소화
-model$tot.withinss
-model$totss
-model$tot.withinss/model$totss
-ospredict <- cl_predict(model, norm_osGAX)
-norm_osGAX
-head(ospredict)
-ospredict
-GAX <- GOLD.AX
-GAX <- GAX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-lagret <- (GAX - Lag(GAX,20))/Lag(GAX,20)
-direction[lagret >0.02]<-"Up"
-direction[lagret < -0.02]<-"Down"
-direction[lagret < 0.02 & lagret > -0.02]<-"Nowhere"
-isdir <- direction[isrow]
-osdir <- direction[osrow]
-neighborhood <- 3
-set.seed(1)
-dim(norm_isGAX)
-dim(norm_osGAX)
-model <- knn(norm_isGAX, norm_osGAX, isdir, neighborhood)
-model
-head(model)
-summary(model)
-matrix <- confusionMatrix(table(model, osdir))
-matrix
-diag(matrix$table)
-#for 사용해 confusionmatrix의 행렬 계산, 총 대각선 요소-총대각선 요소 수/총데이터요소수
-accuracy<- NULL
-for(i in c(1:100)){
-model <- knn(isGAX, osGAX, isdir, i)
-matrix <- confusionMatrix(table(model,osdir))
-diag <- sum(diag(matrix$table))
-total <- sum(matrix$table)
-accuracy[i] <- (total - diag)/total
+plot(ROCperf)
+data("GermanCredit")
+FraudData <- GermanCredit[,1:10]
+head(FraudData)
+#
+len <- dim(FraudData)[1]
+train <- sample(1:len, 0.8*len)
+TrainData <- FraudData[train,]
+TestData <- FraudData[-train,]
+library(randomForest)
+fraud_model <- randomForest(Class ~. , data=TrainData, ntree =50, proximity = TRUE)
+print(fraud_model)
+plot(fraud_model)
+plot(fraud_model)
+importance(fraud_model)
+#
+TestPred <- predict(fraud_model, newdata = TestData)
+table(TestPred, TestData$Class)
+library(PerformanceAnalytics)
+data(edhec)
+data <- edhec["1999", 3:5]
+data
+colnames(data) = c("DC", "EM", "EMN")
+data
+#
+wts <- xtx(matrix(c(0.3, 0.3, 0.4), nrow =1, ncol = 3), as.Date("1998-12-31"))
+#
+wts <- wtx(matrix(c(0.3, 0.3, 0.4), nrow =1, ncol = 3), as.Date("1998-12-31"))
+#
+wts <- xts(matrix(c(0.3, 0.3, 0.4), nrow =1, ncol = 3), as.Date("1998-12-31"))
+colnames(wts) <- colnames(data)
+wts
+#
+Return.portfolio(data, weights = wts, rebalance_on = "months", verbose =TRUE)
+library(randomForest)
+library(mlbench)
+library(caret)
+data("Shuttle")
+Analsis_Data <- head(Shuttle, 10000)
+Analysis_Data <- head(Shuttle, 10000)
+X <- Analysis_Data
+X <- Analysis_Data[,1:9]
+Y <- Analysis_Data[,10]
+X <- Analysis_Data[,1:9] #~9열
+Y <- Analysis_Data[,10] #10열
+control <- trainControl(method = "repeatedcv", number =5, repeats = 3)
+seed <- 4
+metric <- "Accuracy"
+set.seed(seed)
+Count_var <- sqrt(ncol(X))
+tunegrid <- expand.grid(.mtry = Count_var)
+set.seed(seed)
+Count_var <- sqrt(ncol(X))
+tunegrid <- expand.grid(.mtry = Count_var)
+rf_baseline <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+print(rf_baseline)
+plot(rf_gridsearch_method)
+#
+control <- trainControl(method = "repeatedcv", number =5, repeats = 3, search= "grid")
+set.seed(seed)
+tunegrid <- expand.grid(.mtry = c(1:8))
+rf_gridsearch_method <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+set.seed(seed)
+tunegrid <- expand.grid(.mtry = c(1:8))
+rf_gridsearch_method <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+#
+control <- trainControl(method = "repeatedcv", number =5, repeats = 3, search= "grid")
+set.seed(seed)
+tunegrid <- expand.grid(.mtry = c(1:8))
+rf_gridsearch_method <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+#유전자알고리즘
+library(genalg)
+library(ggplot2)
+data("Shuttle")
+Analysis_Data <- head(Shuttle, 10000)
+X <- Analysis_Data[,1:9] #~9열
+Y <- Analysis_Data[,10] #10열
+control <- trainControl(method = "repeatedcv", number =5, repeats = 3)
+seed <- 4
+metric <- "Accuracy"
+set.seed(seed)
+Count_var <- sqrt(ncol(X))
+tunegrid <- expand.grid(.mtry = Count_var)
+rf_baseline <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+print(rf_baseline)
+# better tool
+control <- trainControl(method = "repeatedcv", number =5, repeats = 3, search= "grid")
+set.seed(seed)
+tunegrid <- expand.grid(.mtry = c(1:8))
+rf_gridsearch_method <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+print(rf_gridsearch_method)
+plot(rf_gridsearch_method)
+plot(rf_gridsearch_method)
+View(rf_gridsearch_method)
+plot(rf_gridsearch_method
+plot(rf_gridsearch_method)
+plot(rf_gridsearch_method)
+plot(data)
+plot(rf_gridsearch_method)
+print(rf_gridsearch_method)
+plot(rf_gridsearch_method)
+source('C:/Users/Shinhyunjin/Dropbox/금융공학/R Programming/Learning Quantitative Finance with R/Ch8 Optimization.R', encoding = 'UTF-8', echo=TRUE)
+InputDataset <- data.frame(Stocks = c("Stock1", "Stock2", "Stock3","Stock4","Stock5", "Stock6"), retruns = c(10,11,15,20,12,13), weight = c(0.1, 0.2, 0.1, 0.2, 0.2, 0.3))
+WTlimit <- 1
+InputDataset
+#
+evaluationFunc <- function(x){
+current_solution_returns <- x %*% InputDataset$retruns
+current_solution_weight <- x %*% InputDataset$weight
+if(current_solution_weight > WTlimit)
+return(0) else return(-current_solution_returns)
+}
+#
+GAmodel <- rbga.bin(size =6, popSize =100, iters =50, mutationChange = 0.01, elitism = T, evalFunc = evaluationFunc)
+cat(summary(GAmodel))
+#
+GAmodel <- rbga.bin(size =6, popSize =100, iters =50, mutationChance = 0.01, elitism = T, evalFunc = evaluationFunc)
+cat(summary(GAmodel))
+install.packages("GA")
+library(GA)
+install.packages("GA")
+install.packages("GA")
+library(PerformanceAnalytics)
+library(randomForest)
+library(mlbench)
+library(caret)
+#유전자알고리즘
+library(genalg)
+library(ggplot2)
+install.packages("GA")
+library(GA)
+InputDataset <- data.frame(Stocks = c("Stock1", "Stock2", "Stock3","Stock4","Stock5", "Stock6"), retruns = c(10,11,15,20,12,13), weight = c(0.1, 0.2, 0.1, 0.2, 0.2, 0.3))
+WTlimit <- 1
+InputDataset
+#
+evaluationFunc <- function(x){
+current_solution_returns <- x %*% InputDataset$retruns
+current_solution_weight <- x %*% InputDataset$weight
+if(current_solution_weight > WTlimit)
+return(0) else return(-current_solution_returns)
+}
+#
+GAmodel <- rbga.bin(size =6, popSize =100, iters =50, mutationChance = 0.01, elitism = T, evalFunc = evaluationFunc)
+cat(summary(GAmodel)) #GA result에서 제외되어야할 주식의 비중을 말해준다.
+#
+#
+data(economics)
+Data_Analysis <- data.frame(economics[,2:4])
+head(Data_Analysis)
 }
-accuracy
-plot(accuracy,type = "l")
-plot(accuracy,label = "Accuracy of K % " ,type = "l")
-#for 사용해 confusionmatrix의 행렬 계산, 총 대각선 요소-총대각선 요소 수/총데이터요소수
-accuracy<- NULL
-for(i in c(1:100)){
-model <- knn(isGAX, osGAX, isdir, i)
-matrix <- confusionMatrix(table(model,osdir))
-diag <- sum(diag(matrix$table))
-total <- sum(matrix$table)
-accuracy[i] <- diag/total
+return(SSE)
+OLS_GA <- funciton(Data_Analysis, a0, a1, a2){
+attach(Data_Analysis, warn.conflicts = F)
+Y_hat <- a0 + a1*pop + a2*psavert
+SSE = t(pce-Y_hat) %*% (pce-Y_har)
+detach(Data_Analysis)
+return(SSE)
+}
+#
+OLS_GA <- funciton(Data_Analysis, a0, a1, a2){
+attach(Data_Analysis, warn.conflicts = F)
+Y_hat <- a0 + a1*pop + a2*psavert
+SSE = t(pce-Y_hat) %*% (pce-Y_har)
+detach(Data_Analysis)
+return(SSE)
+}
+#
+OLS_GA <- funciton(Data_Analysis, a0, a1, a2){
+attach(Data_Analysis, warn.conflicts = F)
+Y_hat <- a0 + a1*pop + a2*psavert
+SSE = t(pce-Y_hat) %*% (pce-Y_har)
+detach(Data_Analysis)
+return(SSE)
+}
+detach(Data_Analysis)
+OLS_GA <- funciton(Data_Analysis, a0, a1, a2){
+attach(Data_Analysis, warn.conflicts = F)
+Y_hat <- a0 + a1*pop + a2*psavert
+SSE = t(pce-Y_hat) %*% (pce-Y_hat)
+detach(Data_Analysis)
+return(SSE)
+}
+Data_Analysis
+#
+OLS_GA <- function(Data_Analysis, a0, a1, a2){
+attach(Data_Analysis, warn.conflicts = F)
+Y_hat <- a0 + a1*pop + a2*psavert
+SSE = t(pce-Y_hat) %*% (pce-Y_hat)
+detach(Data_Analysis)
+return(SSE)
+}
+#
+ga.OLS_GA <- ga(type = 'real-valued', min= c(-100,-100,-100), max = c(100,100,100), popSize=500, maxiter = 500, names=c('intercept', 'pop','psavert'), keepBest=T, fitness = function(a) - OLS_GA(Data_Analysis, a[1],a[2],a[3]))
+summary(ga.OLS_GA)
+install.packages("fOptions")
+library(fOptions)
+model <- GBSOption(TypeFlag = "c", S = 900, X = 950, Time = 1/4, r =0.02, sigma =0.22, b= 0.02)
+# type : c or p, 기초자산, 행사가격, 만기, 무위험이자율, 변동성, 보유비용순
+model
+GBSOption(TypeFlag = "p", S = 900, X = 950, Time = 1/4, r = 0.02, sigma = 0.22, b = 0.02)
+CRRBinomialTreeOption(TypeFlag = "ce", S =900, X =950, Time = 1/4, r = 0.02, b = 0.02, sigma = 0.22, n = 3)
+CRRBinomialTreeOption(TypeFlag = "pe", S = 900, X = 950, Time = 1/4, r = 0.02, b = 0.02, sigma = 0.22, n = 3)
+#
+model <- BinomialTreeOption(TypeFlag = "ce", S = 900, X = 950, Time  = 1/4, r = 0.02, b = 0.02, sigma =0.22, n=3)
+#
+model <- BinomialTreeOption(TypeFlag = "ce", S = 900, X = 950, Time  = 1/4, r = 0.02, b = 0.02, sigma =0.22, n=3)
+BinomialTreePlot(model, dy =1, xlab = "Time steps", ylab = "Option Value", xlim = c(0,4), ylim=c(-3,4))
+title(main  = "Call Option Tree")
+retrun(pr)}
+retrun(pr)}
+return(pr)}
+return(pr)}
+##함수로 정의
+func <- function(n){
+pr <- CRRBinomialTreeOption(TypeFlag = "ce", S = 900, X = 950, Time = 1/4, r = 0.02, b =0.02, sigma = 0.22, n = n)@price
+return(pr)
 }
-accuracy
-plot(accuracy ,type = "l")
-install.packages('tidyverse')
-### ##K Cluster 모형 CumRet 구하는법
-GAX <- GOLD.AX
-GAx <- GOLD.AX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-signal <- ifelse( model== "Up", 1, ifelse( model== "Down",-1,0))
-testrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-ret <- GAX/lag(GAX,1) -1
-ret <- ret[testrow]
-ret
-cost <- 0
-length(signal)
-length(ret)
-length(testrow)
-trade_ret <- ret * Lag(signal) - cost
-signal
-length(signal)
-dim(ret)
-cumm_ret <- Return.cumulative(trade_ret)
-anual_ret <- Return.annualized(trade_ret)
-charts.PerformanceSummary(trade_ret)
-GAX <- GOLD.AX
-GAx <- GOLD.AX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-GAX <- GOLD.AX
-GAX <- GOLD.AX[,"GOLD.AX.Close"]
-GAX <- na.omit(GAX)
-signal <- ifelse( model== "Up", 1, ifelse( model== "Down",-1,0))
-testrow <- which(index(GAX) >= ossd & index(GAX) <= osed)
-ret <- GAX/lag(GAX,1) -1
-ret <- ret[testrow]
-ret
-cost <- 0
-length(signal)
-length(ret)
-length(testrow)
-trade_ret <- ret * Lag(signal) - cost
-signal
-length(signal)
-dim(ret)
-cumm_ret <- Return.cumulative(trade_ret)
-anual_ret <- Return.annualized(trade_ret)
-charts.PerformanceSummary(trade_ret)
-plot(GAX)
+#
+price <- sapply(1:100, func) # 1~100반
+#
+price <- sapply(1:100, func) # 1~100반
+#
+price <- sapply(1:100, func) # 1~100반복
+plot(price, type="l", xlab - "Number of steps", ylab = "Option Value")
+plot(price, type="l", xlab = "Number of steps", ylab = "Option Value")
+bs_price <- GBSOption(TypeFlag = "c", S = 900, X = 950, Time = 1/4, r = 0.02, sigma = 0.22, b = 0.02)@price
+abline(h = bs_price, col = 'red')
+legend("topright", legend = c('CRR-price', 'BS-price'), col = c('black', 'red'), pch = 19)
+title(main = "Call Option Pricing models")
+GBSGreeks(Selection = "delta", TypeFlag = "c", S = 900, X = 950, Time = 1/4, r = 0.02, b = 0.02, sigma = 0.22)
+GBSGreeks(Selection = "gamma", TypeFlag = "c", S = 900, X = 950, Time = 1/4, r = 0.02, b = 0.02, sigma  =0.22)
+#
+portfolio <- sapply(c('c', 'p'), function(otype){sapply(500:1500, function(price){
+GBSGreeks(Selection = 'delta',
+TypeFlag = otype,
+S=price, X = 950,
+Time = 1/4, r = 0.02,
+b = 0.02,
+sigma =0.22)
+})
+})
+head(protfolio)
+head(portfolio)
+## Straddle Delta
+plot(500:1500, rowSums(portfolio), type = 'l', xlab = 'underlying Price', ylab = 'Straddle Delta')
+install.packages("RQuantLib", type = 'binary')
+library(RQuantLib)
+iv <- EuropeanOptionImpliedVolatility("call", 11.10, 100, 100, 0.01, 0.03 ,05, 0.4)
+iv
+iv_a <- AmericanOptionImpliedVolatility("call", 11.10, 100, 100, 0.01, 0.03 ,05, 0.4)
+iv_a
+library(fOptions)
+install.packages("RQuantLib", type = 'binary')
+library(RQuantLib)
+install.packages("termstrc")
+library(termstrc)
+install.packages("termstrc")
+library(termstrc)
+install.packages("termstrc")
+library(termstrc)
+install.packages("termstrc")
+library(termstrc)
+library(termstrc)
+library(termstrc)
+install.packages("termstrc")
+library(termstrc)
+library(termstrc)
+data(govbonds)
+install.packages("termstrc
+")
+install.packages("termstrc", lib = "/data/Rpackges/")
+install.packages("termstrc", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/")
+library(termstrc)
+install.packages("~/R/win-library/3.4/termstrc.zip", repos = NULL, type = "win.binary")
+install.packages("termstrc", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("termstrc", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("termstrc-master", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("termstrc-master", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("termstrc", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+library(termstrc, lib.loc= "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("termstrc.zip", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("termstrc", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("~/R/win-library/3.4/termstrc_1.3.7.tar.gz", repos = NULL, type = "source")
+library(termstrc)
+install.packages("~/R/win-library/3.4/termstrc.tar.gz", repos = NULL, type = "source")
+install.packages("termstrc", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+library(termstrc)
+install.packages("termstrc", lib = "C:/Users/Shinhyunjin/Documents/R/win-library/3.4/")
+install.packages("https://cran.r-project.org/src/contrib/Archive/termstrc/")
+install.packages("termstrc", repos="http://R-Forge.R-project.org")
+library(termstrc)
+library(rgl)
+library(termstrc)
+library(termstrc.tar.gz)
+install.packages("termstrc", repos="http://R-Forge.R-project.org")
+install.packages("termstrc", repos="http://R-Forge.R-project.org")
+install.packages("CreditMetrics")
+library(CreditMetrics)
+rc <- c("AAA", "AA", "A", "BBB","BB","B", "CCC", "D")
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creidmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+raw.data
+cm.cs(raw.data, lgd)
+raw.data
+type(raw.data)
+dtype(raw.data)
+raw.data <- data.frame(raw.data)
+lgd <- 0.2 # 부도시손실률
+cm.cs(raw.data, lgd)
+raw.data <- data.frame(raw.data)
+raw.data
+raw.data <- matrix(c(raw.data))
+raw.data
+lgd <- 0.2 # 부도시손실률
+cm.cs(raw.data, lgd)
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+raw.data <- data.frame(raw.data)
+### 9-2 신용파생상품  ###
+rc <- c("AAA", "AA", "A", "BBB", "BB", "B", "CCC", "Default")
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+raw.data <- data.frame(raw.data)
+raw.data <- matrix(c(raw.data), 8,8,dimnames = list(rc,rc), byrow=TRUE)
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+#raw.data <- data.frame(raw.data)
+raw.data <- matrix(c(raw.data), 8,8,dimnames = list(rc,rc), byrow=TRUE)
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+View(raw.data)
+raw.data
+### 9-2 신용파생상품  ###
+rc <- c("AAA", "AA", "A", "BBB", "BB", "B", "CCC", "Default")
+#raw.data <- data.frame(raw.data)
+raw.data <- matrix(c(raw.data), 8,8,dimnames = list(rc,rc), byrow=TRUE)
+raw.data
+### 9-2 신용파생상품  ###
+rc <- c("AAA", "AA", "A", "BBB", "BB", "B", "CCC", "Default")
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+#raw.data <- data.frame(raw.data)
+raw.data <- matrix(c(raw.data), dimnames = list(rc,rc), byrow=TRUE)
+#raw.data <- data.frame(raw.data)
+raw.data <- matrix(c(raw.data), dimnames = list(rc,rc))
+### 9-2 신용파생상품  ###
+rc <- c("AAA", "AA", "A", "BBB", "BB", "B", "CCC", "Default")
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T, fileEncoding="UTF-8-BOM")
+raw.data
+raw.data[1,1]
+raw.data <- raw.data[,1:8]
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T , fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[,2:9]
+raw.data
+raw.data <- matrix(c(raw.data), dimnames = list(rc,rc))
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T , fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[1:8,2:9]
+raw.data
+raw.data
+raw.data <- matrix(c(raw.data), 8,8,dimnames = list(rc,rc))
+raw.data
+lgd <- 0.2 # 부도시손실률
+cm.cs(raw.data, lgd)
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T , fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[1:8,2:9]
+raw.data
+raw.data <- matrix(c(raw.data), dimnames = list(rc,rc))
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T , fileEncoding="UTF-8-BOM")
+raw.data
+raw.data <- raw.data[1:8,2:9]
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T , fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[0:8,2:9]
+raw.data
+raw.data <- matrix(c(raw.data))
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T , fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[2:9,2:9]
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', header=T , fileEncoding="UTF-8-BOM")
+raw.data
+View(raw.data)
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[1:8,2:9]
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', fileEncoding="UTF-8-BOM")
+raw.data
+View(raw.data)
+raw.data <- raw.data[],1:9]
+raw.data <- raw.data[,1:9]
+raw.data
+View(raw.data)
+raw.data <- raw.data[2:8, 2:8]
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[1:8, 1:8]
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[1:8, 2:9]
+raw.data
+raw.data <- matrix(raw.data, nrwo = 8, ncol = 8, byrow = TRUE, dimnames = list(rc,rc))
+raw.data <- matrix(raw.data, nrow = 8, ncol = 8, byrow = TRUE, dimnames = list(rc,rc))
+raw.data
+lgd <- 0.2 # 부도시손실률
+cm.cs(raw.data, lgd)
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv', fileEncoding="UTF-8-BOM")
+raw.data <- raw.data[1:8, 2:9]
+#raw.data <- raw.data[1:8, 2:9]
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv',header= F fileEncoding="UTF-8-BOM")
+#raw.data <- raw.data[1:8, 2:9]
+raw.data
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv',header= F ,fileEncoding="UTF-8-BOM")
+#raw.data <- raw.data[1:8, 2:9]
+raw.data
+raw.data <- matrix(raw.data, nrow = 8, ncol = 8, byrow = TRUE, dimnames = list(rc,rc))
+raw.data
+raw.data <- matrix(c(raw.data), nrow = 8, ncol = 8, byrow = TRUE, dimnames = list(rc,rc))
+raw.data
+lgd <- 0.2 # 부도시손실률
+cm.cs(raw.data, lgd)
+#raw.data <- raw.data[1:8, 2:9]
+raw.data
+### 9-2 신용파생상품  ###
+rc <- c("AAA", "AA", "A", "BBB", "BB", "B", "CCC", "Default")
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv',header= F ,fileEncoding="UTF-8-BOM")
+raw.data
+c(raw.data)
+raw.data <- matrix(raw.data, nrow = 8, ncol = 8, byrow = TRUE, dimnames = list(rc,rc))
+raw.data <- read.csv('C:/Users/Shinhyunjin/Dropbox/data/creditmatrix.csv',header= F ,fileEncoding="UTF-8-BOM")
+#raw.data <- raw.data[1:8, 2:9]
+raw.data
+raw.data <- matrix(raw.data, nrow = 8, ncol = 8, byrow = TRUE, dimnames = list(rc,rc))
+raw.data
+M <- matrix(c(90.81, 8.33, 0.68, 0.06, 0.08, 0.02, 0.01, 0.01,
+0.70, 90.65, 7.79, 0.64, 0.06, 0.13, 0.02, 0.01,
+0.09, 2.27, 91.05, 5.52, 0.74, 0.26, 0.01, 0.06,
+0.02, 0.33, 5.95, 85.93, 5.30, 1.17, 1.12, 0.18,
+0.03, 0.14, 0.67, 7.73, 80.53, 8.84, 1.00, 1.06,
+0.01, 0.11, 0.24, 0.43, 6.48, 83.46, 4.07, 5.20,
+0.21, 0, 0.22, 1.30, 2.38, 11.24, 64.86, 19.79,
+0, 0, 0, 0, 0, 0, 0, 100
+)/100, 8, 8, dimnames = list(rc, rc), byrow = TRUE)
+lgd <- 0.2 # 부도시손실률
+cm.cs(raw.data, lgd)
+cm.cs(M, lgd)
+#
+ead <- c(140000,100000,100000) # 부도시 exposure
+N <- 3 # firms
+n <- 50000 #난수
+r <- 0.03 #rf
+rating <- c("BBB", "AA", "B")
+firmnames <- c("Blizzard", "Activision", "Nexon")
+alpha <- 0.99
+rho <- matrix(c(1,0.4,0.6,0.4,1,0.5,0.6,0.5,1),3,3,dimnames=list(firmnames, firmnames),byrow=TRUE)
+cm.CVaR(M, lgd, ead, N, n,r,rho,alpha,rating)
+pnl <- cm.gain(M, lgd, ead, N, n, r, rho,rating)
+pnl
+install.packages("credule")
+library(credule) #신용파생상
+yct = c(1,2,3,4,5,7)
+ycr = c(0.0050, 0.0070, 0.0080, 0.0100, 0.0120, 0.0150)
+cct = c(1,3,5,7)
+ccsp = c(0.99, 0.98, 0.95, 0.92)
+tenors = c(1,3,5,7)
+yct = c(1,2,3,4,5,7) #테너 (테너 : 채무발생일 ~ 만기일간 기한, 결제기간이라고함)
+ycr = c(0.0050, 0.0070, 0.0080, 0.0100, 0.0120, 0.0150) #수익률곡선할인률
+cct = c(1,3,5,7) #테너
+ccsp = c(0.99, 0.98, 0.95, 0.92) #생존확률
+tenors = c(1,3,5,7) #만기
+r = 0.4 #회수율
+priceCDS(yct, ycr, cct, ccsp, tenor, r)
+priceCDS(yct, ycr, cct, ccsp, tenors, r)
+#
+cdsSpreads = c(0.0050, 0.0070, 0.0090, 0.0110)
+bootstrapCDS(yct,ycr,cct,ccsp,r)
+install.packages("GUIDE")
+library(GUIDE)
+## 금리파생상품 ##
+irswapvalue()
+## 금리파생상품 ##
+irswapvalue()
+install.packages("fExoticOptions")
+library(fExoticOptions)
+price <- GeometricAverageRateOption("c", 110, 120, 0.5, 0.03, 0.05 ,0.1)
+price
+#2번 방법
+TurnbullWakemanAsianApproxOption(TypeFlag = "p", S = 100, SA = 102, X = 120, Time = 0.5,
+time = 0.25, tau = 0, r = 0.03, b = 0.05, sigma = 0.1)@price
+#3번 방법
+LevyAsianApproxOption(TypeFlag = "p", S = 100, SA = 102, X = 120, Time = 0.5,
+time = 0.25, r = 0.03, b = 0.05, sigma = 0.1)
+#3번 방법
+LevyAsianApproxOption(TypeFlag = "p", S = 100, SA = 102, X = 120, Time = 0.5,
+time = 0.25, r = 0.03, b = 0.05, sigma = 0.1)@price
+# Down and Out
+StandardBarrierOption(TypeFlag = "cdo", S  =100, X=90, H = 95, K = 3, Time=0.5, r = 0.08, b = 0.04, sigma = 0.25)@price
+#2번 Up and out down and out call (Double Barrier)
+DoubleBarrierOption(TypeFlag = "co", S = 100,X=100, L = 50, U = 150, TIme =0.25,
+r = 0.1 ,b=0.1, sigma = 0.15, delta1 = -0.1, delta2 = 0.1)@price
+#2번 Up and out down and out call (Double Barrier)
+DoubleBarrierOption(TypeFlag = "co", S = 100,X=100, L = 50, U = 150, Time =0.25,
+r = 0.1 ,b=0.1, sigma = 0.15, delta1 = -0.1, delta2 = 0.1)@price
+#3번 룩백 Barrier Up and out
+LookBarrierOption(TypeFlag = "cuo", S = 100, X = 100, H = 130, Time1 = 0.25, Time2 = 1, r=0.1,b=0.1, sigma=0.15)@price
+#3번 룩백 Barrier Up and out
+LookBarrierOption(TypeFlag = "cuo", S = 100, X = 100, H = 130, time1 = 0.25, Time2 = 1, r=0.1,b=0.1, sigma=0.15)@price
+#4번 Gap Digital
+GapOption(TypeFlag = "c", S = 50, X1= 50, X2= 57, Time = 0.5, r = 0.09, b = 0.09, sigma = 0.20)
+## 5번 Cash or Nothing : 만기시점에 기초자산가격이 행사가격에 도달할 경우 미리 정한가 지불 ##
+CashOrNothingOption(TypeFlag = "p", S = 100, X = 80, K = 10, Time = 9/12, r = 0.06, b = 0, sigma = 0.35)
+TwoAssetCashOrNothingOption(TypeFlag = "c", S1 = 100, S2 = 100, X1 = 110, X2 = 90,
+K = 10, Time = 0.5, r=0.1, b1 = 0.05, b2= 0.06, sigma1 = 0.2, sigma2=0.25,
+rho = 0.5)@price
diff --git a/Ch1 Basic R.R b/Ch1 Basic R.R
new file mode 100644
index 0000000..b3e97e5
--- /dev/null
+++ b/Ch1 Basic R.R	
@@ -0,0 +1,197 @@
+### Ch1. R 속으로 ###
+
+## 패키지 설치 ##
+#자동설치법 
+install.packages("ggplot2")
+install.packages("forecast")
+library(ggplot2)
+library(forecast)
+#수동설치법 
+install.packages("ggplot2", lib = "/data/Rpackages/")
+library(ggplot2, lib.loc = "/data/Rpackages/")
+
+## Data ##
+#class(변수)는 변수의 데이터 형태를 말해준다.
+# Character
+a <- "hello"
+print(class(a))
+# numeric
+b <- 2.5
+print(class(b))
+# integer
+c <- 6L
+print(class(c))
+#complex
+d <- 1 + 2i
+print(class(d))
+# Logic 
+e<- FALSE
+print(class(e))
+
+## Data 유형 ##
+#1. Vector
+
+a <- "Quantitative"
+b <- "Finance"
+c(a,b)
+
+v <- c(1,2,3)
+v
+
+#2. List
+
+list1 = list(c(4,5,6), "hello" ,24.5)
+print(list1)
+print(list1[2]) #두번째 요소 
+
+#list merge
+list1 <- list(5,6,7) #3행 
+list2 <- list("a","b","c") #3행 
+combined_list <- c(list1, list2)
+print(combined_list) #6행 
+
+#3.Matrix
+m <- matrix(c(1,2,3,4,5,6), nrow =2, ncol =3) #nrow행, ncol열 
+print(m)
+
+#4. array
+a <- array(c(4,5), dim = c(3,3,2)) #c(행,열,차원) 
+print(a)
+
+#5. Factor
+
+a <- c(2,3,4,2,3)
+fact <- factor(a) #서로 다른 값을 구분해 저장 
+print(fact) 
+print(nlevels(fact)) #서로 다른 값의 개수 
+
+#6. Data Frame
+
+data <- data.frame(
+  Name = c("A","B","c"), age = c(18,20,23), gender = c("F","M","M")
+)
+print(data)
+
+
+## 데이터 내보내고 갖고 오기 import and export ##
+
+print(getwd()) #디렉토리 경로 
+#setwd는 디렉토리를 설정해준다.
+
+#csv
+data <- read.csv("C:/Users/Shinhyunjin/Dropbox/data/gold.csv")
+data
+print(is.data.frame(data)) #dataframe을 확인
+
+write.csv(data,"result.csv") #csv 쓰기 
+output <- read.csv("result.csv")
+print(output)
+
+#xlsx
+install.packages('rJava')
+library(rJava)
+install.packages("xlsx", INSTALL_opts=c("--no-multiarch"))
+any(grepl("xlsx", installed.packages())) #설치확인 
+library(xlsx)
+xlsxdata <- read.xlsx("C:/Users/Shinhyunjin/Dropbox/data/hw1data.xlsx", sheetIndex=1)
+
+output <- write.xlsx(data, "result.xlsx")
+output <- read.csv("result.csv")
+print(output)
+
+#Web data
+
+URL <- "http://finance.google.co.uk/finance/historical?q=GOOGL
++&startdate=JAN+01%2c+2016
++&enddate=DEC+31%2c+2016
++&output=csv"
+google<- as.data.frame(read.csv(URL))
+head(google)
+
+#Database 
+install.packages("RMySQL")
+library(RMySQL)
+
+## 함수 ##
+#value까지 순차적으로 제곱을 반복적으로 도출하는 함수 
+findingSqrFunc <- function(value){
+  for(j in 1:value){
+    sqr <- j^2
+    print(sqr)
+  }
+}
+
+findingSqrFunc(4)
+
+#인수없이 5의 배수 도출하는 함수
+
+function_test <- function(){
+  for(i in 1:3){
+    print(i*5)
+  }
+}
+function_test()
+
+#인수가 있는 함수
+function_test2 <- function(a,b,c){
+  result<-a*b+c
+  print(result)
+}
+function_test2(1,2,3)
+
+### 반복문 ###
+
+#1. if
+x <-5
+if(x>0){
+  print("참")
+} else{
+  print("거짓")
+}
+
+#2. for 정의된 갯수만큼 실행 
+var <- c(3,6,8,9,11,16)
+counter <- 0
+for (val in var){
+  if(val%%2 !=0) counter=counter+1 #홀수를 세는 코드 -> 2의 배수가 아니다!
+}
+print(counter)
+
+#3. While 계속 반복함
+var <- c("hello")
+counter <- 4
+while(counter < 7){
+  print(var)
+  counter = counter+1
+}
+
+#4. apply
+
+sample = matrix(c(1:10), nrow =5, ncol = 2) #5행 2열 
+apply(sample, 1, sum) # 데이터, 행별(2이면 열별) , 합 의미
+
+#5. sapply
+sapply(1:5, function(x) x^3)
+
+#6. break
+
+vec <- c("hello")
+counter <- 5
+repeat {
+  print(vec)
+  counter <- counter +1
+  if(counter>8){
+    break
+  }
+}
+
+#7. next
+vec<- c(2,3,4,5,6)
+for (i in vec){
+  if(i==4){
+    next # 4에서 건너뛰는데 이때 4를 제외하고 건너뜀 (next)
+  }
+  print(i)
+} 
+  }
+}
\ No newline at end of file
diff --git a/Ch1. Basic R b/Ch1. Basic R
new file mode 100644
index 0000000..86c4c3a
--- /dev/null
+++ b/Ch1. Basic R	
@@ -0,0 +1,140 @@
+### Ch1. R 속으로 ###
+
+## 패키지 설치 ##
+#자동설치법 
+install.packages("ggplot2")
+install.packages("forecast")
+library(ggplot2)
+library(forecast)
+#수동설치법 
+install.packages("ggplot2", lib = "/data/Rpackages/")
+library(ggplot2, lib.loc = "/data/Rpackages/")
+
+## Data ##
+#class(변수)는 변수의 데이터 형태를 말해준다.
+# Character
+a <- "hello"
+print(class(a))
+# numeric
+b <- 2.5
+print(class(b))
+# integer
+c <- 6L
+print(class(c))
+#complex
+d <- 1 + 2i
+print(class(d))
+# Logic 
+e<- FALSE
+print(class(e))
+
+## Data 유형 ##
+#1. Vector
+
+a <- "Quantitative"
+b <- "Finance"
+c(a,b)
+
+v <- c(1,2,3)
+v
+
+#2. List
+
+list1 = list(c(4,5,6), "hello" ,24.5)
+print(list1)
+print(list1[2]) #두번째 요소 
+
+#list merge
+list1 <- list(5,6,7) #3행 
+list2 <- list("a","b","c") #3행 
+combined_list <- c(list1, list2)
+print(combined_list) #6행 
+
+#3.Matrix
+m <- matrix(c(1,2,3,4,5,6), nrow =2, ncol =3) #nrow행, ncol열 
+print(m)
+
+#4. array
+a <- array(c(4,5), dim = c(3,3,2)) #c(행,열,차원) 
+print(a)
+
+#5. Factor
+
+a <- c(2,3,4,2,3)
+fact <- factor(a) #서로 다른 값을 구분해 저장 
+print(fact) 
+print(nlevels(fact)) #서로 다른 값의 개수 
+
+#6. Data Frame
+
+data <- data.frame(
+  Name = c("A","B","c"), age = c(18,20,23), gender = c("F","M","M")
+)
+print(data)
+
+
+## 데이터 내보내고 갖고 오기 import and export ##
+
+print(getwd()) #디렉토리 경로 
+#setwd는 디렉토리를 설정해준다.
+
+#csv
+data <- read.csv("C:/Users/Shinhyunjin/Dropbox/data/gold.csv")
+data
+print(is.data.frame(data)) #dataframe을 확인
+
+write.csv(data,"result.csv") #csv 쓰기 
+output <- read.csv("result.csv")
+print(output)
+
+#xlsx
+install.packages('rJava')
+library(rJava)
+install.packages("xlsx", INSTALL_opts=c("--no-multiarch"))
+any(grepl("xlsx", installed.packages())) #설치확인 
+library(xlsx)
+xlsxdata <- read.xlsx("C:/Users/Shinhyunjin/Dropbox/data/hw1data.xlsx", sheetIndex=1)
+
+output <- write.xlsx(data, "result.xlsx")
+output <- read.csv("result.csv")
+print(output)
+
+#Web data
+
+URL <- "http://finance.google.co.uk/finance/historical?q=GOOGL
++&startdate=JAN+01%2c+2016
++&enddate=DEC+31%2c+2016
++&output=csv"
+google<- as.data.frame(read.csv(URL))
+head(google)
+
+#Database 
+install.packages("RMySQL")
+library(RMySQL)
+
+## 함수 ##
+#value까지 순차적으로 제곱을 반복적으로 도출하는 함수 
+findingSqrFunc <- function(value){
+  for(j in 1:value){
+    sqr <- j^2
+    print(sqr)
+  }
+}
+
+findingSqrFunc(4)
+
+#인수없이 5의 배수 도출하는 함수
+
+function_test <- function(){
+  for(i in 1:3){
+    print(i*5)
+  }
+}
+function_test()
+
+#인수가 있는 함수
+function_test2 <- function(a,b,c){
+  result<-a*b+c
+  print(result)
+}
+function_test2(1,2,3)
\ No newline at end of file
diff --git a/Ch1. Basic R Programming b/Ch1. Basic R Programming
new file mode 100644
index 0000000..a4cce2b
--- /dev/null
+++ b/Ch1. Basic R Programming	
@@ -0,0 +1 @@
+### Ch1. Basic R Programming ##
diff --git a/Ch1.R b/Ch1.R
new file mode 100644
index 0000000..b3e97e5
--- /dev/null
+++ b/Ch1.R
@@ -0,0 +1,197 @@
+### Ch1. R 속으로 ###
+
+## 패키지 설치 ##
+#자동설치법 
+install.packages("ggplot2")
+install.packages("forecast")
+library(ggplot2)
+library(forecast)
+#수동설치법 
+install.packages("ggplot2", lib = "/data/Rpackages/")
+library(ggplot2, lib.loc = "/data/Rpackages/")
+
+## Data ##
+#class(변수)는 변수의 데이터 형태를 말해준다.
+# Character
+a <- "hello"
+print(class(a))
+# numeric
+b <- 2.5
+print(class(b))
+# integer
+c <- 6L
+print(class(c))
+#complex
+d <- 1 + 2i
+print(class(d))
+# Logic 
+e<- FALSE
+print(class(e))
+
+## Data 유형 ##
+#1. Vector
+
+a <- "Quantitative"
+b <- "Finance"
+c(a,b)
+
+v <- c(1,2,3)
+v
+
+#2. List
+
+list1 = list(c(4,5,6), "hello" ,24.5)
+print(list1)
+print(list1[2]) #두번째 요소 
+
+#list merge
+list1 <- list(5,6,7) #3행 
+list2 <- list("a","b","c") #3행 
+combined_list <- c(list1, list2)
+print(combined_list) #6행 
+
+#3.Matrix
+m <- matrix(c(1,2,3,4,5,6), nrow =2, ncol =3) #nrow행, ncol열 
+print(m)
+
+#4. array
+a <- array(c(4,5), dim = c(3,3,2)) #c(행,열,차원) 
+print(a)
+
+#5. Factor
+
+a <- c(2,3,4,2,3)
+fact <- factor(a) #서로 다른 값을 구분해 저장 
+print(fact) 
+print(nlevels(fact)) #서로 다른 값의 개수 
+
+#6. Data Frame
+
+data <- data.frame(
+  Name = c("A","B","c"), age = c(18,20,23), gender = c("F","M","M")
+)
+print(data)
+
+
+## 데이터 내보내고 갖고 오기 import and export ##
+
+print(getwd()) #디렉토리 경로 
+#setwd는 디렉토리를 설정해준다.
+
+#csv
+data <- read.csv("C:/Users/Shinhyunjin/Dropbox/data/gold.csv")
+data
+print(is.data.frame(data)) #dataframe을 확인
+
+write.csv(data,"result.csv") #csv 쓰기 
+output <- read.csv("result.csv")
+print(output)
+
+#xlsx
+install.packages('rJava')
+library(rJava)
+install.packages("xlsx", INSTALL_opts=c("--no-multiarch"))
+any(grepl("xlsx", installed.packages())) #설치확인 
+library(xlsx)
+xlsxdata <- read.xlsx("C:/Users/Shinhyunjin/Dropbox/data/hw1data.xlsx", sheetIndex=1)
+
+output <- write.xlsx(data, "result.xlsx")
+output <- read.csv("result.csv")
+print(output)
+
+#Web data
+
+URL <- "http://finance.google.co.uk/finance/historical?q=GOOGL
++&startdate=JAN+01%2c+2016
++&enddate=DEC+31%2c+2016
++&output=csv"
+google<- as.data.frame(read.csv(URL))
+head(google)
+
+#Database 
+install.packages("RMySQL")
+library(RMySQL)
+
+## 함수 ##
+#value까지 순차적으로 제곱을 반복적으로 도출하는 함수 
+findingSqrFunc <- function(value){
+  for(j in 1:value){
+    sqr <- j^2
+    print(sqr)
+  }
+}
+
+findingSqrFunc(4)
+
+#인수없이 5의 배수 도출하는 함수
+
+function_test <- function(){
+  for(i in 1:3){
+    print(i*5)
+  }
+}
+function_test()
+
+#인수가 있는 함수
+function_test2 <- function(a,b,c){
+  result<-a*b+c
+  print(result)
+}
+function_test2(1,2,3)
+
+### 반복문 ###
+
+#1. if
+x <-5
+if(x>0){
+  print("참")
+} else{
+  print("거짓")
+}
+
+#2. for 정의된 갯수만큼 실행 
+var <- c(3,6,8,9,11,16)
+counter <- 0
+for (val in var){
+  if(val%%2 !=0) counter=counter+1 #홀수를 세는 코드 -> 2의 배수가 아니다!
+}
+print(counter)
+
+#3. While 계속 반복함
+var <- c("hello")
+counter <- 4
+while(counter < 7){
+  print(var)
+  counter = counter+1
+}
+
+#4. apply
+
+sample = matrix(c(1:10), nrow =5, ncol = 2) #5행 2열 
+apply(sample, 1, sum) # 데이터, 행별(2이면 열별) , 합 의미
+
+#5. sapply
+sapply(1:5, function(x) x^3)
+
+#6. break
+
+vec <- c("hello")
+counter <- 5
+repeat {
+  print(vec)
+  counter <- counter +1
+  if(counter>8){
+    break
+  }
+}
+
+#7. next
+vec<- c(2,3,4,5,6)
+for (i in vec){
+  if(i==4){
+    next # 4에서 건너뛰는데 이때 4를 제외하고 건너뜀 (next)
+  }
+  print(i)
+} 
+  }
+}
\ No newline at end of file
diff --git a/Ch2 Basic Statistics.R b/Ch2 Basic Statistics.R
new file mode 100644
index 0000000..8a31e18
--- /dev/null
+++ b/Ch2 Basic Statistics.R	
@@ -0,0 +1,257 @@
+### Ch2. Basic Statistics : Modeling ###
+
+
+### 1. Probability Distribution ###
+
+#1. Normal Distribution #
+
+Sampledata = read.csv("C:/Users/Shinhyunjin/Dropbox/data/WTI etf.csv")
+head(Sampledata)
+
+#누적밀도함수 dnorm
+y <- dnorm(Sampledata$price, mean = mean(Sampledata$price), sd = sd(Sampledata$price, na.rm= FALSE))
+plot(Sampledata$price, y)
+
+#누적분포함수 pnorm
+pnorm(.02, mean= mean(Sampledata$price), sd = sd(Sampledata$price, na.rm=FALSE))
+
+#분위수함수 qnorm
+qnorm(0.159837, mean= mean(Sampledata$price), sd=sd(Sampledata$price, na.rm=FALSE),lower.tail=FALSE)
+
+#난수행성함수 rnorm
+rnorm(5, mean=mean(Sampledata$price), sd = sd(Sampledata$price, na.rm =FALSE))
+
+#2. Lognormal Distribution
+
+#dlnorm 로그정규분포의 밀도함수
+
+y <- dlnorm(Sampledata$volume, meanlog = mean(Sampledata$volume), sd=sd(Sampledata$volume, na.rm=FALSE)
+            )
+plot(Sampledata$volume, y)
+
+#plnorm 로그정규분포의 누적확률분포함수
+y <- plnorm(Sampledata$volume, meanlog = mean(Sampledata$volume), sdlog = sd(Sampledata$volume, na.r=FALSE))
+plot(Sampledata$volume,y)
+
+#qlnorm 로그정규분포의 q분위수 구하기
+#rlnorm 로그정ㅂ분포의 난수생성 
+
+
+##3. 포아송 ##
+
+ppois(15, lambda = 10)
+ppois(15, lambda = 10, lower= FALSE) #오른쪽 꼬리 확률 
+
+##4. 균등분포 이론- 연속균등분포 ##
+
+runif(10, min = 1, max= 5)
+
+##5. 극단값 이론 ##
+
+install.packages("POT") # peaks over threshold by generalized pareto distribution
+library(POT)
+#
+data(ardieres)
+abc <- ardieres[1:10000,]
+events <- clust(abc,u=1.5, tim.cond=8/365, clust.max= TRUE)
+par(mfrow = c(2,2))
+mrlplot(events[,"obs"])
+diplot(events)
+tcplot(events[,"obs"], which = 1)
+tcplot(events[,"obs"], which = 2)
+#
+obs <- events[,"obs"]
+ModelFit <- fitgpd(obs, thresh= 5, "pwmu")
+ModelFit
+
+
+### 3. 표본 추출 ###
+
+##3-1 무작위 표본추출 ##
+#비복원 -> 중복을 불가하게한 무작위표본 추출 
+RandomSample <- Sampledata[sample(1:nrow(Sampledata), 10, replace = FALSE),]
+RandomSample
+#복원 -> 중복을 허용한 무작위표본 추출 
+RandomSample <- Sampledata[sample(1:nrow(Sampledata), 10, replace = TRUE),]
+RandomSample
+
+## 3-2 층화 표본추출 ##
+
+install.packages("sampling")
+library(sampling)
+table(Sampledata$volume, Sampledata$price)
+
+#다른 그룹에서 표본 추출
+
+Stratsubset = strata(Sampledata, c("volume", "price"), size = c(2,2,2,2), method="srswor")
+Stratsubset
+
+### 4. 통계량 ###
+
+#평균값
+mean(Sampledata$price)
+#중앙값
+median(Sampledata$price)
+#최빈값
+findmode <- function(x){
+  uniqx <- unique(x)
+  uniqx[which.max(tabulate(match(x, uniqx)))]
+}
+findmode(Sampledata$price)
+#요약
+summary(Sampledata$price)
+#적률
+library(e1071)
+moment(Sampledata$price, order =3, center=TRUE)
+#첨도 -> 뾰족한 정도 
+kurtosis(Sampledata$price)
+#왜도 -> 분포의 대칭성, 평균이 중앙갑보다 작으면 left-skewed, 크면 right-skewed.
+skewness(Sampledata$price) #크다 -> right-skewed네
+
+### 5. 상관관계 ###
+
+install.packages("Hmisc")
+library(Hmisc)
+
+x <- Sampledata[,2:3]
+x<-na.omit(x)
+rcorr(as.matrix(x), type = "pearson")
+acf(Sampledata$price)
+pacf(Sampledata$price)
+ccf(Sampledata$price, Sampledata$volume, main = "ccf plot")
+
+### 6. 가설검정 ###
+
+## 6-1 분산을 아는 모평균의 왼쪽꼬리검정 ##
+xbar = 9.9
+mu0 = 10
+sig = 1.1
+n = 30
+z= (xbar-mu0)/(sig/sqrt(n))
+z
+
+alpha = 0.05
+z.alpha = qnorm(1-alpha)
+-z.alpha
+
+pnorm(z) #0.05보다 크므로 채택 
+
+## 6-2 분산을 아는 모평균의 오른쪽꼬리검정 ##
+
+xbar = 5.1
+mu0 = 5
+sig = 0.25
+n = 30
+z = (xbar-mu0)/(sig/sqrt(n))
+z
+alpha=0.5
+z.alpha = qnorm(1-alpha)
+z.alpha
+pnorm(z, lower.tail=FALSE) #0.05보다 작으므로 기각 
+
+## 6-3 분산을 아는 모평균의 양측검정 ##
+
+xbar = 1.5
+mu0 = 2
+sig = 0.1
+n = 30
+z= (xbar-mu0)/(sig/sqrt(n))
+z
+
+alpha=0.05
+z.half.alpha=qnorm(1-alpha/2)
+c(-z.half.alpha, z.half.alpha)
+2*pnorm(z) # 0.05보다 작으므로 귀무가설을 기각한다. 
+
+## 6-4. 분산을 모르는 t 모평균의 왼쪽꼬리검정 
+
+xbar = 0.9
+mu0 = 1
+sig = 0.1
+n = 30
+t = (xbar-mu0)/(sig/sqrt(n))
+
+alpha= 0.05
+t.alpha = qt(1-alpha, df= n-1)
+-t.alpha
+
+pt(t, df=n-1) # 0.05보다 작으므로 귀무가설을 기각한다 .
+
+## 6-5. 분산을 모르는 t 모평균의 오른쪽꼬리검정
+
+xbar = 3.1
+mu0 = 3
+sig = 0.2
+n = 30
+t = (xbar-mu0)/(sig/sqrt(n))
+t
+alpha= 0.05
+t.alpha = qt(1-alpha, df= n-1)
+t.alpha
+
+alpha = 0.05
+t.alpha = qt(1-alpha, df = n-1)
+t.alpha
+pt(t, df = n-1, lower.tail = FALSE) #0.05보다 작으므로 귀무가설을 기각한다 .
+
+## 6-6. 분산을 모르는 모평균의 양측검정 \
+
+xbar = 1.9
+mu0 = 2
+sig = 0.1
+n = 30
+t = (xbar-mu0)/(sig/sqrt(n))
+t
+
+alpha= 0.05
+t.half.alpha = qt(1-alpha, df= n-1)
+c(-t.half.alpha, t.half.alpha)
+
+### 7. 파라미터 추정 ###
+
+## 7-1 MLE (Maximum Likelihood Estimation ##) ##
+install.packages("stats4")
+library(stats4)
+set.seed(100)
+NO_values <- 100
+Y <- rnorm(NO_values, mean =5, sd =1)
+mean(Y)
+sd(Y)
+LogL <- function(mu, sigma){
+  A = dnorm(Y, mu, sigma)
+  - sum(log(A))
+}
+
+mle(LogL, start = list(mu=2, sigma=2), method = "L-BFGS-B", lower = c(-Inf, 0), upper=c(Inf,Inf))
+
+## 7-2 선형모델 
+
+Y <- Sampledata$price
+X <- Sampledata$volume
+fit <- lm (Y~X)
+summary(fit)
+
+### 8 이상치검출 ###
+
+## 8-1 상자 
+boxplot(Sampledata$volume, main = "Volume", boxwex=0.1)
+
+## 8-2 Local outlier factor dkfrhflwma
+
+install.packages("DMwR")
+library(DMwR)
+outlier.scores <- lofactor(Sampledata$volume, k =4)
+plot(density(outlier.scores))
+order(outlier.scores, decreasing = T)[1:5] #상위 5개의 이상치 , 행번호가 출력됨 
+
+### 9. 표준화와 정규화 ###
+
+## 9-1 표준화 
+scale(Sampledata$volume, center =TRUE, scale =FALSE) # 중심화
+scale(Sampledata$volume, center = TRUE, scale =TRUE) # 표준화 
+
+## 9-2 정규화
+
+normalized = (Sampledata$volume-min(Sampledata$volume))/
+  (max(Sampledata$volume)-min(Sampledata$volume))
+normalized
diff --git a/Ch2.R b/Ch2.R
new file mode 100644
index 0000000..e63f844
--- /dev/null
+++ b/Ch2.R
@@ -0,0 +1,3 @@
+### Ch2. Basic Statistics ###
+
+# 
diff --git a/Ch3 Statistical Analysis.R b/Ch3 Statistical Analysis.R
new file mode 100644
index 0000000..2208f28
--- /dev/null
+++ b/Ch3 Statistical Analysis.R	
@@ -0,0 +1,167 @@
+### Ch3. Statistical Analysis ###
+
+
+### 1. Linear Regression ##
+
+data <- read.csv("C:/Users/Shinhyunjin/Dropbox/data/rdata1.csv")
+head(data)
+
+#scatter
+Yprice = data$price.etf.autos
+Xprice = data$price.etf.insurance
+scatter.smooth(Yprice, Xprice, type="p",xlab = "ETF insur", ylab = "ETF autos")
+
+#regression
+LinearR.lm = lm(Yprice ~ Xprice, data = data)
+coeffs = coefficients(LinearR.lm);coeffs
+
+predict(LinearR.lm)
+
+summary(LinearR.lm)$r.squared
+summary(LinearR.lm)
+
+#선형회귀모델의 신뢰구간
+Predictdata = data.frame(Xprice = 8000)
+predict(LinearR.lm, Predictdata, interval = "confidence") # 예측값 신뢰구간
+
+#잔차도
+LinearR.res = resid(LinearR.lm)
+plot(Xprice, LinearR.res, ylab = "Residuals",xlab = "Xprice",main = "Residual Plot")
+#표준화잔차도
+LinearRSTD.res = rstandard(LinearR.lm)
+plot(Xprice, LinearRSTD.res, ylab = "Standardized Residuals", xlab = "Xprice", main = "Residual Plot")
+
+#오차의 정규분포
+qqnorm(LinearRSTD.res, ylab = "Standardized Residuals", xlab = "Normal Scores", main = "Error Normal Distribution plot")
+qqline(LinearRSTD.res)
+
+### 2. Multiple Regression ###
+
+data2 <- read.csv("C:/Users/Shinhyunjin/Dropbox/data/rdata1.csv")
+Yprice <- data2$etf.kospi
+X1price <- data2$price.etf.autos
+X2price <- data2$price.etf.insurance
+X3price <- data2$price.etf.energy.chemical
+X4price <- data2$price.etf.bank
+MultipleR.lm = lm(Yprice ~ X1price + X2price + X3price + X4price, data = data2)
+summary(MultipleR.lm)
+
+#예측값
+newdata = data.frame(X1price = 13000, X2price = 7900, X3price = 11200, X4price = 7400)
+predict(MultipleR.lm, newdata)
+predict(MultipleR.lm, newdata, interval = "confidence")#confidence level prediction
+
+### 3. Multicollinearity "
+
+install.packages("car")
+library(car)
+vif(MultipleR.lm)
+
+### 4. ANOVA ###
+data3 <- read.csv("C:/Users/Shinhyunjin/Dropbox/data/rdata1.csv")
+Yprice <- data3$etf.kospi
+X1price <- data3$price.etf.autos
+boxplot(Yprice ~ X1price)
+
+oneway.test(data3$etf.kospi ~ data3$price.etf.autos, var.equal = TRUE)
+
+### 4. Feature Selection ###
+#상관관계
+data4 <- read.csv("C:/Users/Shinhyunjin/Dropbox/data/rdata1.csv")
+Yprice <- data4$etf.kospi
+X1price <- data4$price.etf.autos
+X2price <- data4$price.etf.insurance
+X3price <- data4$price.etf.energy.chemical
+X4price <- data4$price.etf.bank
+correlationMatrix <- cor(data4[,2:5])
+correlationMatrix
+
+### 5. 단계식 변수선택 ###
+
+install.packages("MASS")
+library(MASS)
+MultipleR.lm = lm(Yprice ~ X1price+X2price+X3price+X4price, data = data4)
+step <- stepAIC(MultipleR.lm, direction = "both")
+step$anova
+
+#분류에의한 변수선택
+install.packages("mlbench")
+install.packages("caret")
+library(mlbench)
+library(caret)
+library(randomForest)
+
+control <-rfeControl(functions = rfFuncs, method = "cv", number = 10)
+Output <- rfe(data4[,3:6],data4[,2:3], sizes = c(3:6), rfeControl = control)
+predictors(Output)
+plot(Output, type = c("g","o"))
+
+###6. 웨이블릿 분석 ###
+
+install.packages("wavelets")
+library(wavelets)
+library(quantmod)
+
+getSymbols("HSPX", from = "2018-01-01", to = "2018-12-29")
+HSPX
+
+hspx <- HSPX$HSPX.Close
+ret_hspx <- Delt(hspx, k=1)
+par(mfrow = c(2,1))
+plot(xts(hspx), type = "l")
+plot(xts(ret_hspx), type = "l")
+
+head(hspx)
+tail(hspx)
+
+#웨이블릿 변환
+hspx2 <- as.ts(hspx)
+model <- wavelets::dwt(hspx2, filter = "la8", n.levels = 3)
+model
+model@W #웨이블릿계수 
+model@V #척도화계수
+plot(model)
+
+#하르필터
+model2 <- wavelets::dwt(hspx2, filter = "haar", n.levels = 3)
+plot(model2)
+
+#역이산 웨이블릿
+imodel <- idwt(model2, fast = TRUE)
+
+#다중해상도분석
+model3 <- mra(hspx2, filter = "la8", n.levels = 3)
+
+#최대중첩이산웨이블릿(MODWT)
+model4 <- modwt(hspx2, filter = "la8",n.levels = 5)
+plot.modwt(model4)
+
+### 7. 고속푸리에변환 (FFT) ###
+
+model5 <- fft(hspx2)
+rp = Re(model5) #실수부real part
+ip = Im(model5) #허수부imaginary part
+absmodel <- abs(model5) #절대값
+plot(absmodel)
+
+norm_absmodel <- absmodel[1:(length(hspx2)/2)]
+Angle = atan2(ip,rp)
+
+spec_density <- spectrum(hspx2, method = c("pgram", "ar"))
+
+### 8. 힐버트 변환 ###
+
+install.packages("seewave", repos="http://cran.at.r-project.org/")수
+library(seewave)
+model <- hilbert(hspx2,1) #1은 주파수
+summary(model)
+
+rp<-Re(model)
+ip <-Im(model)
+ifreq(hspx2, 1, ylim = c(0,0.0001))
+ifreq(hspx2, 1, phase = "TRUE", ylim=  c(-0.5,1)) #위상을 준다 phase
+phase_diff <- phase1 - phase2
+#값도출
+output = ifreq(hspx2, 1, plot =FALSE)
+freq <- output$f
+phase <- output$p
diff --git a/Ch4 time series analysis.R b/Ch4 time series analysis.R
new file mode 100644
index 0000000..ba52d0f
--- /dev/null
+++ b/Ch4 time series analysis.R	
@@ -0,0 +1,60 @@
+### Ch4 Time Series Analysis ###
+  
+### 3. ARCH GARCH 시리즈 ###
+
+## 3-1 GARCH BAsic ##
+install.packages("rugarch")
+library(rugarch)
+
+gspec.ru <- ugarchspec(mean.model = list(armaOrder = c(0,0)),distribution = "std")
+gfit.ru <- ugarchfit(gspec.ru, price)
+coef(gfit.ru)
+
+#forecast
+FutureForecast = ugarchforecast(gfit.ru, n.ahead=5)
+FutureForecast
+
+## 3-2 EGARCH -> 지수형 GARCH로서 시장시나리오에 적합함 ##
+# return form + no N/A data
+return <- na.omit(return)
+egarchetf.spec = ugarchspec(variance.model = list(model = "eGARCH", garchOrder=c(1,1)),mean.model = list(armaOrder = c(0,0)))
+egarchetf.fit = ugarchfit(egarchetf.spec, return)
+egarchetf.fit
+coef(egarchetf.fit)
+
+FutureForecast = ugarchforecast(egarchetf.fit, n.ahead =5)
+FutureForecast
+
+## 3-3 VGARCH -> 벡터 GARCH 및 다변량 GARCH라고 한다.##
+
+install.packages("rmgarch")
+install.packages("PerformanceAnalytics")
+library(rmgarch)
+library(PerformanceAnalytics)
+
+getSymbols("114800.KS", from = "2018-01-01", to = "2018-12-31")
+data1 <- `114800.KS`
+price1 <- data1$`114800.KS.Close`
+return1 <- Delt(price1)
+return1 <- na.omit(return1)
+
+getSymbols("122630.KS", from = "2018-01-01", to = "2018-12-31")
+data2 <- `122630.KS`
+price2 <- data2$`122630.KS.Close`
+return2 <- Delt(price2)
+return2 <- na.omit(return2)
+
+data <- cbind(return1, return2)
+
+garch_spec = ugarchspec(mean.model = list(armaOrder = c(2,1)), variance.model = list(garchOrder = c(1,1),model = "sGARCH"),distribution.model = "norm")
+dcc.garch_spec = dccspec(uspec = multispec(replicate(2, garch_spec)), dccOrder = c(1,1),distribution = "mvnorm")
+dcc_fit = dccfit(dcc.garch_spec, data = data)
+fcst = dccforecast(dcc_fit, n.ahead =5)
+fcst
+
+## 3-4 DCC 동적조건부 상관관계-> 간결하고 설명력 좋음.##
+
+garch_spec2 = ugarchspec(mean.model = list(armaOrder = c(0,0)), variance.model = list(garchOrder = c(1,1),model = "sGARCH"),distribution.model = "norm")
+dcc.garch_spec2 = dccspec(uspec = multispec(replicate(2, garch_spec2)), dccOrder = c(1,1),distribution = "mvnorm")
+dcc_fit2 = dccfit(dcc.garch_spec2, data = data, fit.control = list(scale=TRUE))
+dcc_fit2
diff --git a/Ch5 Algorithm Trading.R b/Ch5 Algorithm Trading.R
new file mode 100644
index 0000000..6487bf4
--- /dev/null
+++ b/Ch5 Algorithm Trading.R	
@@ -0,0 +1,420 @@
+### Ch5 Algorithm Trading ###
+
+library(quantmod)
+library(PerformanceAnalytics)
+library(tseries)
+library(fPortfolio)
+
+## 5-1 모멘텀 / 방향성 트레이딩 ##
+
+getSymbols("226490.KS", from = "2015-09-09", to= "2018-12-31")
+getSymbols("169950.KS", from = "2015-09-09", to = "2018-12-31")
+
+kospietf <- `226490.KS`
+kospietf_p <- kospietf$`226490.KS.Close`
+kospietf_p <- na.omit(kospietf_p)
+
+kospietf_r <- Delt(kospietf_p, k = 1)
+kospietf_r <- na.omit(kospietf_r)
+
+chinaetf <- `169950.KS`
+chinaetf_p <- chinaetf$`169950.KS.Close`
+chinaetf_p <- na.omit(chinaetf_p)
+
+chinaetf_r <- Delt(chinaetf_p, k = 1)
+chinaetf_r <- na.omit(chinaetf_r)
+
+#추세분석
+par(mfrow = c(2,1))
+plot(kospietf_p,type="l")
+plot(kospietf_r, type = "l")
+
+#날짜 나누기
+in_sd <- "2015-09-09"
+in_ed <- "2017-12-31"
+out_sd <- "2018-01-01"
+out_ed <- Sys.Date()
+
+in_kospi <- kospietf_p[(index(kospietf_p) >= in_sd & index(kospietf_p) <= in_ed),]
+in_kospi <- na.omit(in_kospi)
+in_ret_kospit <- kospietf_r[(index(kospietf_r) >= in_sd & index(kospietf_r) <= in_ed),]
+out_kospi <- kospietf_p[(index(kospietf_p) >= out_sd & index(kospietf_p) <= out_ed),]
+out_kospi <- na.omit(out_kospi)
+out_ret_kospit <- kospietf_r[(index(kospietf_r) >= out_sd & index(kospietf_r) <= out_ed),]
+
+
+#Variables
+
+macd <- MACD(in_kospi, nFast=7, nSlow = 12, nSig = 15, maType="SMA", percent = FALSE)
+#macd : 이동평균선 사이의 관계 -> 추세방향과 주가 움직임 분석에 용이
+bb <- BBands(in_kospi, n=20, maType="SMA", sd = 2)
+#bb : 상하한선의 폭 분석, 변동성에 비례함, 과매수매도 분석
+signal <- NULL
+signal <- ifelse(in_kospi > bb[,'up'] & macd[,'macd'] > macd[,'signal'],1,ifelse(in_kospi<bb[,'dn']&macd[,'macd']<macd[,'signal'],-1,0))
+
+#Investment Results
+trade_return <- in_ret_kospit*lag(signal)
+cumm_ret <- Return.cumulative(trade_return)
+annual_ret <- Return.annualized(trade_return)
+
+charts.PerformanceSummary(trade_return)
+summary(as.ts(trade_return))
+
+#Financial Results
+maxDrawdown(trade_return)
+StdDev(trade_return)
+StdDev.annualized(trade_return)
+VaR(trade_return, p = 0.95)
+SharpeRatio(as.ts(trade_return), Rf = 0.018, p = 0.95, FUN = "StdDev")
+SharpeRatio.annualized(trade_return, Rf = 0.018)
+
+## 표본외 데이터 계산 #
+#Variables
+macd_o <- MACD(out_kospi, nFast=7, nSlow = 12, nSig = 15, maType="SMA", percent = FALSE)
+#macd : 이동평균선 사이의 관계 -> 추세방향과 주가 움직임 분석에 용이
+bb_o <- BBands(out_kospi, n=20, maType="SMA", sd = 2)
+#bb : 상하한선의 폭 분석, 변동성에 비례함, 과매수매도 분석
+signal_o <- NULL
+signal_o <- ifelse(out_kospi > bb_o[,'up'] & macd_o[,'macd'] > macd_o[,'signal'],1,ifelse(out_kospi<bb_o[,'dn']&macd_o[,'macd']<macd_o[,'signal'],-1,0))
+
+#Investment Results
+trade_return_o <- out_ret_kospit*lag(signal_o)
+cumm_ret_o <- Return.cumulative(trade_return_o)
+annual_ret_o <- Return.annualized(trade_return_o)
+
+charts.PerformanceSummary(trade_return_o)
+summary(as.ts(trade_return_o))
+
+#Financial Results
+maxDrawdown(trade_return_o)
+StdDev(trade_return_o)
+StdDev.annualized(trade_return_o)
+VaR(trade_return_o, p = 0.95)
+SharpeRatio(as.ts(trade_return_o), Rf = 0.018, p = 0.95, FUN = "StdDev")
+SharpeRatio.annualized(trade_return_o, Rf = 0.018)
+
+## var-cov matrix ##
+
+var(kospietf_r, na.rm = T)
+var(chinaetf_r, na.rm = T)
+var(kospietf_r + chinaetf_r, na.rm =T) #cov때문에 둘이 다를 수밖에 없다.
+
+sd(kospietf_r, na.rm = T)
+sd(chinaetf_rm na.rm = T)
+cor(kospietf_r[!is.na(kospietf_r)], chinaetf_r[!is.na(chinaetf_r)])
+
+pf_ret <- data.frame(matrix(NA, dim(kospietf_r)[1],2))
+pf_ret[,1] <- kospietf_r
+pf_ret[,2] <- chinaetf_r
+pf_ret <- pf_ret[!is.na(pf_ret[,1]),]
+cor(pf_ret)
+
+
+### 5-2 Pair Trading ###
+
+kospietf_r[1] <- 1
+chinaetf_r[1] <- 1
+
+norm_kospi <- apply(kospietf_r, 2, cumsum)
+norm_china <- apply(chinaetf_r, 2, cumsum)
+
+#Basic Results 
+plot(norm_kospi, type="l", ylim = c(0.75,1.35),ylab="Normalized_Price",main="KODEX Kospi ETF & China ETF")
+lines(norm_china, col = "red")
+legend('topright', c("KOSPI_ETF", "CHINA_ETF"), lty = 1, col=c('black', 'red'),bty = 'o', cex=1)
+
+#Advanced Results
+norm_kospi <- xts(norm_kospi, index(kospietf_r))
+norm_china <- xts(norm_china, index(chinaetf_r))
+
+par(mfrow  = c(3,1))
+#plot(norm_kospi, type="l", ylim = c(0.75,1.35),ylab="Normalized_Price",main="KODEX Kospi ETF & China ETF")
+lines(norm_china, col = "red")
+legend('topright', c("KOSPI_ETF", "CHINA_ETF"), lty = 1, col=c('black', 'red'),bty = 'o', cex=1)
+
+diff = norm_kospi - norm_china
+plot(diff, type = "l", ylab = "Normalized_Price_difference")
+
+me <- mean(diff)
+std <- sd(diff)
+n<-1
+
+ub <- me + n*std
+lb <- me - n*std
+signal <- ifelse(diff>ub, 1, ifelse(diff<lb, -1,0))
+me_dynamic <- rollapply(diff, 10, mean)
+std_dynamic <- rollapply(diff, 10, sd)
+plot(signal, type = "l")
+
+cost <- 0
+spread_return <- kospietf_r - chinaetf_r
+trade_return <- spread_return * lag(signal) - cost
+summary(trade_return)
+
+#Final Results
+cumm_ret <- Return.cumulative(trade_return)
+annual_ret <- Return.annualized(trade_return)
+charts.PerformanceSummary(trade_return)
+maxxdd <- maxDrawdown(trade_return)
+sd <- StdDev(trade_return)
+sda <- StdDev.annualized(trade_return)
+VaR(trade_return, p = 0.95)
+SharpeRatio(as.ts(trade_return), Rf = 0, p = 0.95, FUN = "StdDev")
+SharpeRatio.annualized(trade_return, Rf = 0)
+
+### 5-3 상관관게 기반 페어트레이딩 ###
+
+data <- data.frame(matrix(NA, dim(kospietf_r)[1],2))
+data[,1] <- kospietf_r
+data[,2] <- chinaetf_r
+
+data <- xts(data, index(kospietf_r))
+correlation <- function(x){
+  result <- cor(x[,1], x[,2])
+  return(result)
+}
+corr <- rollapply(data,252,correlation, by.column= FALSE)
+plot(corr)
+#
+hedge_ratio <- kospietf_p / chinaetf_p
+
+roll_me <- rollapply(hedge_ratio, 14, mean)
+roll_std <- rollapply(hedge_ratio, 14, sd)
+n <-1
+roll_ub <- roll_me + n*roll_std
+roll_lb <- roll_me - n*roll_std
+
+signal <- NULL
+signal <- ifelse(hedge_ratio > roll_ub, -1, ifelse(hedge_ratio < roll_lb, 1, 0))
+lagsignal <- Lag(signal, 1)
+signal <- ifelse(lagsignal == -1 & hedge_ratio > roll_me,-1,ifelse(lagsignal ==1 & hedge_ratio < roll_me, 1,0))
+
+spread_return <- kospietf_r - chinaetf_r
+trade_return <- spread_return * lag(signal) - cost
+
+### 5-4 공적분 기반 페어 트레이딩 ###
+adf.test(kospietf_p)
+
+diff <- kospietf_p - Lag(kospietf_p, 1) #1차차분 
+adf.test(diff[!is.na(diff)])
+
+model <- lm(kospietf_p ~ chinaetf_p +0)
+model
+summary(model)
+adf.test(as.ts(model$residuals))
+
+plot(kospietf_p, type = "l", main = "KODEX KOSPI ETF & CHINA ETF")
+par(mfrow = c(2,1))
+lines(chinaetf_p * model$coefficients, col = "red")
+plot(as.xts(model$residuals), type = "l")
+
+roll_me <- rollapply(model$residuals, 14, mean)
+roll_std <- rollapply(model$residuals, 14, sd)
+n <- 1
+roll_ub <- roll_me + n * roll_std
+roll_lb <- roll_me - n * roll_std
+signal <- NULL
+signal <- ifelse(model$residuals > roll_ub, -1, ifelse(model$residuals < roll_lb, 1, 0))
+lagsignal <- Lag(signal, 1)
+signal <- ifelse(lagsignal ==-1 & model$residuals > roll_me, -1, ifelse(lagsignal ==1 & model$residuals < roll_me, 1, 0))
+
+#### 5-4 CAPM ####
+pfdata <- cbind(kospietf_r, chinaetf_r)
+
+rf <- rep(0, dim(pfdata)[1])
+model <- lm((pfdata[,2] -rf ) ~ (pfdata[,1]- rf))
+model
+
+CAPM.beta(pfdata[,2], pfdata[,1])
+CAPM.alpha(pfdata[,2], pfdata[,1])
+plot(as.ts(chinaetf_r), as.ts(kospietf_r), xlim = c(0,0.04), ylim = c(0,0.04),xlab = "CHINA ETF Return", ylab="KOSPI ETF Return")
+abline(model, col = "red")
+
+#### 5-5 Multi Factor Model -> do not Run ####
+
+con = gzcon(url('http://www.systematicportfolio.com/sit.gz', 'rb'))
+source(con)
+close(con)
+
+dow.jones.components <- function(){
+  url = 'http://money.cnn.com/data/dow30/'
+  txt = join(readLines(url))
+  temp = gsub(pattern = '">', replacement = '<td>', txt, perl =TRUE)
+  temp = gsub(pattern = '</a>', replacement = '</td>', temp, perl = TRUE)
+  temp = extract.table.from.webpage(temp, 'Volume', has.header=T)
+  trim(temp[,'Company'])
+}
+tickers=dow.jones.components()
+
+#ticker 크로울링 
+data.fund <- new.env()
+temp = paste(iif(nchar(tickers) <=3, 'NYSE:', 'NASDAQ:'), tickers, sep = '')
+for(i in 1:len(tickers)) data.fund[[tickers[i]]] = fund.data(temp[i], 80)
+save(data.fund, file = 'data.fund.Rdata')
+#가격 크로울링
+data <- new.env()
+getSymbols(tickers, src = 'yahoo', from = '1970-01-01', env = data, auto.assign = T)
+for(i in ls(data)) data[[i]] = adjustOHLC(data[[i]], use.Adusted=T)
+save(data,file='data.Rdata')
+#날짜 변수
+date.fund.data <- function(data){
+  quarter.end.date = as.Date(paste(data['quarter end date',],'/1',sep=''),'%Y/%m/%d')
+  quarterly.indicator = data['quarterly indicator',]
+  date.preliminary.data.loaded = as.Date(data['date preliminary data loaded',],],'%Y-%m-%d')+1
+  months = seq(quarter.end.date[1], tail(quarter.end.date, 1)+365, by = '1 month')
+  index = match(quarter.end.date, months)
+  quarter.end.date = months[iif(quarterly.indicator =='4', index +3, index+2)+1]-1
+  fund.date = date.preliminary.data.loaded
+  fund.date[is.na(fund.date)] = quarter.end.date[is.na(fund.date)]
+  return(fund.date)
+}
+
+## fundamental analysis ##
+
+for(i in tickers){
+  fund = data.fund[[i]]
+  fund.date = date.fund.data(fund)
+  EPS = get.fund.data('Diluted EPS from Total Operations', fund, fund.date, is.12m.rolling= T)
+  CSHO = get.fund.data('total common shares out', fund, fund.date) # common shares outstanding
+  CEQ = get.fund.data('total equity', fund, fund.date) #common equity
+  data[[i]] = merge(data[[i]], EPS, CSHO, CEQ)
+}
+#
+bt.prep(data, align = 'keep.all', dates='1995::2011')
+prices= data$prices
+prices = bt.apply.matrix(prices, function(x) ifna.prev(x))
+#
+factors = list()
+#재무비율
+factors$TV = list()
+#시가총액
+CSHO = bt.apply(data, function(x), ifna.prev(x[,'CSHO']))
+MKVAL = prices * CSHO
+# EPS
+EPS = bt.apply(data, function(x) ifna.prev(x[,'EPS']))
+factors$TV$EP = EPS / prices
+# 자기자본대 주가비율
+CEQ = bt.apply(data, function(x) ifna.prev(x[,'CEQ']))
+factors$TV$BP = CEQ / MKVAL
+
+#횡단면 Z점수로 변환
+for(i in names(factors$TV)){
+  factors$TV[[i]] = (factors$TV[[i]] - cap.wewighted.mean(factors$TV[[i]], MKVAL))/apply(factors$TV[[i]], 1, sd, na.rm =T)
+}
+
+load.packages("abind")
+temp  = abind(factors$TV, along = 3)
+
+factors$TV$AVG = factors$TV[[1]]
+factors$TV$AVG[] = apply(temp, c(1,2), mean, na.rm= T)
+
+#월말 찾기
+month.ends = endpoints(prices, 'months')
+prices = prices[month.ends, ]
+n = ncol (prices)
+nperiods = nrow(prices)
+
+ret = prices / mlag(prices) -1
+next.month.ret = mlag(ret,-1)
+MKVAL = MKVAL[month.ends,]
+for(j in 1:len(factors)){
+  for(i in 1:len(factors[[j]])){
+    factors[[j]][[i]] = factors[[j]][[i]][month.ends,]
+  }
+}
+
+out = compute.quantiles(factors$TV$AVG, next.month.ret, plot = F)
+models=list()
+for(i in 1:5){
+  data$weight[] = NA
+  data$weight[month.ends,] = iif(out$qunatiles ==i, out$weights,0)
+  capital = 100000
+  data$weight[] = (capital/prices) * (data$weight)
+  models[[paste('Q',i,sep='')]] = bt.run(data,type='share', captial = capital)
+}
+
+#스프레드 
+data$weight[] = NA
+data$weight[month.ends,] = iif(out$qunatiles ==5, out$weights, iif(out$qunatiles==1, -out$weights, 0))
+capital = 10000
+data$weight[]=(capital/prices)*(data$weight)
+models$Q5_Q1 = bt.run(data, type = 'share', capital = capital)
+
+factors.avg = list()
+for(j in names(factors)) factors.avg[[j]] = factors[[j]]$AVG
+factors.avg = add.avg.factor(factors.avg)
+nperiods = nrow(nex.month.ret)
+n = ncol(next.month.ret)
+
+#각 요인에 대한 행렬 생성
+factors.matrix = abind(factors.avg, along = 3)
+all.data = factors.matrix
+
+#베타
+beta = all.data[,1,] * NA
+
+# all.data에 next.month.ret추가
+all.data = abind(next.month.ret, all.data, along = 3)
+dimnames(all.data)[[3]][1] = 'Ret'
+all.data[is.na(all.data)] <- 0
+#베타예측(요인수익률)
+for(t in 30:(nperiods-1)){
+  temp = all.data[t:t,,]
+  x = temp[,-1]
+  y = temp[,1]
+  beta[(t+1),]  = lm(y~x-1)$coefficients
+}
+
+#알파수익률 예측 생성
+alpha = next.month.ret * NA
+for(t in 40:(nperiods-1)){
+  coef = colMeans(beta[(t-5):t,], na.rm=T)
+  alpha[t,] = rowSums(all.data[t,,-1] * t(repmat(coef, 1, n)), na.rm = T)
+}
+
+#### 5-5 Portfolio Theory ####
+
+stockData <- new.env()
+symbols <- c("278530.KS","069500.KS","114800.KS","122630.KS")
+start_date <- as.Date("2018-01-01")
+getSymbols(symbols, src= "yahoo", env = stockData, from =start_date)
+x<-list()
+#
+for (i in 1:length(symbols)){
+  x[[i]] <- get(symbols[i], pos = stockData)
+  x[[i]]$gl <- ((Cl(x[[i]]) - Op(x[[i]])) / Op(x[[i]]))*100
+  if(i==1)
+    data<-Cl(x[[i]])
+  else
+    data <- cbind(data, Cl(x[[i]]))
+}
+#
+data_ret <- apply(data,2,Delt)
+data_ret <- na.omit(data_ret)
+napos <- which(apply(data_ret,2,is.na))
+avg_ret <- apply(data_ret,2,mean)
+covariance_mat <- cov(data_ret, use = 'na')
+weights <- c(0.2, 0.3,0.35,0.15)
+#
+source("C:/Users/Shinhyunjin/Dropbox/data/portfolio.R")
+# Normal
+weightedport = getPortfolio(er=avg_ret, cov.mat = covariance_mat, weights =weights)
+weightedport
+# MVP
+minvar_port <- globalMin.portfolio(avg_ret, covariance_mat)
+minvar_port
+# Efficient
+rf <-0
+efficient_port <- efficient.portfolio(avg_ret, covariance_mat, rf)
+efficient_port
+# Tangent
+tangency_port <- tangency.portfolio(avg_ret, covariance_mat, rf)
+tangency_port
+
+## EF
+efficient_frontier <- efficient.frontier(avg_ret, covariance_mat, alpha.min =-2,alpha.max=2, nport=50)
+plot(efficient_frontier, plot.assets =T)
+points(minvar_port$sd, minvar_port$er, col = "blue")
+points(tangency_port$sd, tangency_port$er, col = "red")
+tangent_sharpe_ratio = (tangency_port$er - rf) / tangency_port$sd
+abline(a=rf, b=tangent_sharpe_ratio)
\ No newline at end of file
diff --git a/Ch7 Risk Management.R b/Ch7 Risk Management.R
new file mode 100644
index 0000000..41ba62c
--- /dev/null
+++ b/Ch7 Risk Management.R	
@@ -0,0 +1,204 @@
+#### Ch7. Risk Management ####
+
+library(quantmod)
+library(tseries)
+library(PerformanceAnalytics)
+install.packages("SACCR") #바젤규제 패키지
+library(SACCR)
+library(caret)
+library(caTools)
+library(ROCR)
+library(randomForest)
+### 7-1 시장리스크 ###
+## Beta로 측정 ##
+#Data -> KOSDAQ ETF와 KOSPI ETF
+getSymbols("232080.KS", from = "2016-01-01")
+getSymbols("277630.KS", from = "2016-01-01")
+#Variables
+KOSDA <- `232080.KS`
+KOSPI <- `277630.KS`
+KOSDA <- KOSDA$`232080.KS.Close`
+KOSPI <- KOSPI$`277630.KS.Close`
+KOSDA <- na.omit(KOSDA)
+KOSPI <- na.omit(KOSPI)
+KOSDA_m <- to.monthly(KOSDA)[,"KOSDA.Close"]
+KOSPI_m <- to.monthly(KOSPI)[,"KOSPI.Close"]
+KOSDA_r_m <- Delt(KOSDA_m)
+KOSPI_r_m <- Delt(K
+                  OSPI_m)
+
+#기간
+KOSDA_r_m <- KOSDA_r_m[(index(KOSDA_r_m)>="2017-10")]
+KOSPI_r_m <- KOSPI_r_m[(index(KOSPI_r_m)>="2017-10")]
+
+#결과
+betafit <- lm(KOSPI_r_m ~ KOSDA_r_m)
+result <- summary(betafit)
+result
+result$coefficients[2,1]
+
+### 7-2 포트폴리오 리스크 ###
+
+getSymbols("099140.KS")
+getSymbols("117680.KS")
+getSymbols("226490.KS")
+getSymbols("140710.KS")
+
+China <- `099140.KS`
+Steel <- `117680.KS`
+KOSPI <- `226490.KS`
+Trans <- `140710.KS`
+
+China <- na.omit(China)
+Steel <- na.omit(Steel)
+KOSPI <- na.omit(KOSPI)
+Trans <- na.omit(Trans)
+
+China_m <- to.monthly(China)[,"China.Close"]
+Steel_m <- to.monthly(Steel)[,"Steel.Close"]
+KOSPI_m <- to.monthly(KOSPI)[,"KOSPI.Close"]
+Trans_m <- to.monthly(Trans)[,"Trans.Close"]
+
+China_m_r <- Delt(China_m)
+Steel_m_r <- Delt(Steel_m)
+KOSPI_m_r <- Delt(KOSPI_m)
+Trans_m_r <- Delt(Trans_m)
+
+China_m_r <- China_m_r[(index(China_m_r) >= "2016-01")]
+Steel_m_r <- Steel_m_r[(index(Steel_m_r) >= "2016-01")]
+KOSPI_m_r <- KOSPI_m_r[(index(KOSPI_m_r) >= "2016-01")]
+Trans_m_r <- Trans_m_r[(index(Trans_m_r) >= "2016-01")]
+
+China_m_r_p <- China_m_r - 0.0175
+Steel_m_r_p <- Steel_m_r - 0.0175
+KOSPI_m_r_p <- KOSPI_m_r - 0.0175
+Trans_m_r_p <- Trans_m_r - 0.0175
+
+MeanSD <- rbind(cbind("China ETF", mean(China_m_r_p), sd(China_m_r_p)),
+                cbind("Steel ETF", mean(Steel_m_r_p), sd(Steel_m_r_p)),
+                cbind("KOSPI ETF", mean(KOSPI_m_r_p), sd(KOSPI_m_r_p)),
+                cbind("Trans ETF", mean(Trans_m_r_p), sd(Trans_m_r_p)))
+MeanSD
+
+lm1 <- lm(China_m_r_p ~ KOSPI_m_r_p)
+lm2 <- lm(Steel_m_r_p ~ KOSPI_m_r_p)
+lm3 <- lm(Trans_m_r_p ~ KOSPI_m_r_p)
+
+return_avg <- matrix(c((lm1$coefficients[2]*mean(KOSPI_m_r_p)),
+                       (lm2$coefficients[2]*mean(KOSPI_m_r_p)),
+                       (lm3$coefficients[2]*mean(KOSPI_m_r_p))),nrow=1)
+covariance <- cov(Data) # 원래는 Single Index Model대로 해야하나 일단 이렇게함 
+covariance <- matrix(c(covariance), nrow=3)
+sol <- portfolio.optim(x = return_avg, covmat = covariance, shorts = F)
+sol$pw # Weight
+
+
+### 7-3 Value at Risk (VaR) ###
+
+## 7-3-1 파라미터 VaR - 분산공분산법 ##
+
+mean = 2
+sigma = 4
+alpha = 0.05
+Var_paramatic = qnorm(alpha, mean, sigma)
+Var_paramatic
+
+#Expected Short fall
+alpha_z = qnorm(alpha)
+ES_paramatic = mean + sigma * (dnorm(alpha_z)/(1-alpha))
+ES_paramatic
+
+## 7-3-2 역사적 VaR ##
+#Data
+symbollist = c("AMD", "AAPL", "ORCL")
+getSymbols(symbollist, form = "2017-01-01", to = "2019-01-10")
+AMD = AMD[,"AMD.Adjusted", drop = F]
+AAPL = AAPL[,"AAPL.Adjusted", drop = F]
+ORCL = ORCL[,"ORCL.Adjusted", drop = F]
+AMD_return = CalculateReturns(AMD, method = "log")
+AAPL_return = CalculateReturns(AAPL, method = "log")
+ORCL_return = CalculateReturns(ORCL, method = "log")
+AMD_return = AMD_return[-1,]
+AAPL_return =AAPL_return[-1,]
+ORCL_return = ORCL_return[-1,]
+bind_return <- cbind(AMD_return, AAPL_return, ORCL_return)
+head(bind_return)
+# 역사적 VaR
+HVAR <- VaR(bind_return, p=0.95, method = "historical")
+HVAR
+# ES
+HCVAR <- ES(bind_return, p =0.95, method = "historical")
+HCVAR
+# component VaR
+VaR(bind_return, p=0.95, portfolio_method = "component")
+# Margninal VaR
+VaR(bind_return, p=0.95, portfolio_method = "marginal")
+
+### 7-4 몬테카를로 시뮬레이션 ###
+
+Sample_Size <- 2000
+set.seed(2345)
+Z <- rnorm(Sample_Size)
+mean <- 0.2
+sigma <- 0.25
+deltat <- 0.0833333
+returns <- mean*deltat + sigma*Z*sqrt(deltat)
+hist(returns, breaks =50)
+#
+Mean_new <- mean(returns) *12
+Mean_new
+std_new <- sd(returns)*(12)^(0.5)
+std_new
+#
+VaR(returns, p = 0.95, method = "historical")
+
+### 7-5 바젤 ###
+
+CalcEAD(50, 400) # 대체비용, 예상미래익스포져 순
+
+### 7-6  개인 신용리스크 ###
+data("GermanCredit")
+LRData <- GermanCredit[,1:10]
+str(LRData)
+summary(LRData)
+#
+set.seed(100)
+res = sample.split(LRData$Class, 0.6)
+Train_data = subset(LRData, res == TRUE)
+Test_data = subset(LRData, res == FALSE)
+
+lgfit = glm(Class ~. , data = Train_data, family  = "binomial")
+summary(lgfit)
+#유의한 변수 따로
+lgfit = glm(Class ~ Duration + InstallmentRatePercentage + Age, family = "binomial", data = Train_data)
+summary(lgfit)
+#
+Train_data$predicted.risk = predict(lgfit, newdata = Train_data, type  = "response")
+table(Train_data$Class, as.numeric(Train_data$predicted.risk >=0.05))
+#
+pred= prediction(Train_data$predicted.risk, Train_data$Class)
+as.numeric(performance(pred, "auc")@y.values)
+# 그래프
+predict_Train = predict(lgfit, type = "response")
+ROCpred = prediction(predict_Train, Train_data$Class)
+ROCperf = performance(ROCpred, "tpr", "fpr")
+plot(ROCperf)
+
+### 7-7 사기탐지 ###
+
+data("GermanCredit")
+FraudData <- GermanCredit[,1:10]
+head(FraudData)
+#
+len <- dim(FraudData)[1]
+train <- sample(1:len, 0.8*len)
+TrainData <- FraudData[train,]
+TestData <- FraudData[-train,]
+fraud_model <- randomForest(Class ~. , data=TrainData, ntree =50, proximity = TRUE)
+# 결과
+print(fraud_model)
+plot(fraud_model)
+importance(fraud_model)
+#
+TestPred <- predict(fraud_model, newdata = TestData)
+table(TestPred, TestData$Class)
diff --git a/Ch8 Optimization.R b/Ch8 Optimization.R
new file mode 100644
index 0000000..8c913aa
--- /dev/null
+++ b/Ch8 Optimization.R	
@@ -0,0 +1,80 @@
+### Ch8 Optimization ###
+
+library(PerformanceAnalytics)
+library(randomForest)
+library(mlbench)
+library(caret)
+#유전자알고리즘
+library(genalg)
+library(ggplot2)
+install.packages("GA")
+library(GA)
+### 8-1 주기적 재조정 (dynamic reblancing) ###
+
+data(edhec)
+data <- edhec["1999", 3:5] #3~5 열 
+colnames(data) = c("DC", "EM", "EMN")
+data
+#
+wts <- xts(matrix(c(0.3, 0.3, 0.4), nrow =1, ncol = 3), as.Date("1998-12-31"))
+colnames(wts) <- colnames(data)
+wts
+### Dynamic 조정 ###
+Return.portfolio(data, weights = wts, rebalance_on = "months", verbose =TRUE)
+
+### 8-2 전진분석 walk forward testing ###
+## 최적 파라미터 결정 ##
+
+# 그리드 평가 #
+
+data("Shuttle")
+Analysis_Data <- head(Shuttle, 10000)
+X <- Analysis_Data[,1:9] #~9열
+Y <- Analysis_Data[,10] #10열
+control <- trainControl(method = "repeatedcv", number =5, repeats = 3)
+seed <- 4
+metric <- "Accuracy"
+set.seed(seed)
+Count_var <- sqrt(ncol(X))
+tunegrid <- expand.grid(.mtry = Count_var)
+rf_baseline <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+print(rf_baseline)
+# better tool
+control <- trainControl(method = "repeatedcv", number =5, repeats = 3, search= "grid")
+set.seed(seed)
+tunegrid <- expand.grid(.mtry = c(1:8))
+rf_gridsearch_method <- train(Class ~ ., data = Analysis_Data, method = "rf", metric = metric, tuneGrid = tunegrid, trControl = control)
+print(rf_gridsearch_method)
+plot(rf_gridsearch_method)
+
+
+### 8-3 유전자 알고리즘 ###
+
+InputDataset <- data.frame(Stocks = c("Stock1", "Stock2", "Stock3","Stock4","Stock5", "Stock6"), retruns = c(10,11,15,20,12,13), weight = c(0.1, 0.2, 0.1, 0.2, 0.2, 0.3))
+WTlimit <- 1
+InputDataset
+#
+evaluationFunc <- function(x){
+  current_solution_returns <- x %*% InputDataset$retruns
+  current_solution_weight <- x %*% InputDataset$weight
+  if(current_solution_weight > WTlimit)
+    return(0) else return(-current_solution_returns)
+}
+#
+GAmodel <- rbga.bin(size =6, popSize =100, iters =50, mutationChance = 0.01, elitism = T, evalFunc = evaluationFunc)
+cat(summary(GAmodel)) #GA result에서 제외되어야할 주식의 비중을 말해준다.
+#
+data(economics)
+Data_Analysis <- data.frame(economics[,2:4])
+head(Data_Analysis)
+#
+OLS_GA <- function(Data_Analysis, a0, a1, a2){
+  attach(Data_Analysis, warn.conflicts = F)
+  Y_hat <- a0 + a1*pop + a2*psavert
+  SSE = t(pce-Y_hat) %*% (pce-Y_hat)
+  detach(Data_Analysis)
+  return(SSE)
+}
+#
+ga.OLS_GA <- ga(type = 'real-valued', min= c(-100,-100,-100), max = c(100,100,100), popSize=500, maxiter = 500, names=c('intercept', 'pop','psavert'), keepBest=T, fitness = function(a) - OLS_GA(Data_Analysis, a[1],a[2],a[3]))
+summary(ga.OLS_GA) #intercpet, pop, psavert 계수가 결과값!
diff --git a/Ch9 Derivative Pricing.R b/Ch9 Derivative Pricing.R
new file mode 100644
index 0000000..e08d843
--- /dev/null
+++ b/Ch9 Derivative Pricing.R	
@@ -0,0 +1,154 @@
+##### Ch9. 파생상품 가격결정 #####
+
+library(fOptions)
+install.packages("RQuantLib", type = 'binary')
+library(RQuantLib)
+install.packages("CreditMetrics")
+library(CreditMetrics) #신용파생상품
+install.packages("credule")
+library(credule) #신용파생상품
+install.packages("GUIDE") #금리파생상품
+library(GUIDE)
+install.packages("fExoticOptions")
+library(fExoticOptions)
+### 9-1 바닐라 옵션 ###
+
+## 9-1-1 블랙숄즈 ##
+
+GBSOption(TypeFlag = "c", S = 900, X = 950, Time = 1/4, r =0.02, sigma =0.22, b= 0.02)
+GBSOption(TypeFlag = "p", S = 900, X = 950, Time = 1/4, r = 0.02, sigma = 0.22, b = 0.02)
+# type : c or p, 기초자산, 행사가격, 만기, 무위험이자율, 변동성, 보유비용순
+
+## 9-1-2 Cox Ross Rubinstein -> 이항모형 ##
+
+CRRBinomialTreeOption(TypeFlag = "ce", S =900, X =950, Time = 1/4, r = 0.02, b = 0.02, sigma = 0.22, n = 3)
+CRRBinomialTreeOption(TypeFlag = "pe", S = 900, X = 950, Time = 1/4, r = 0.02, b = 0.02, sigma = 0.22, n = 3)
+#
+model <- BinomialTreeOption(TypeFlag = "ce", S = 900, X = 950, Time  = 1/4, r = 0.02, b = 0.02, sigma =0.22, n=3)
+BinomialTreePlot(model, dy =1, xlab = "Time steps", ylab = "Option Value", xlim = c(0,4), ylim=c(-3,4))
+title(main  = "Call Option Tree")
+##함수로 정의
+func <- function(n){
+  pr <- CRRBinomialTreeOption(TypeFlag = "ce", S = 900, X = 950, Time = 1/4, r = 0.02, b =0.02, sigma = 0.22, n = n)@price
+  return(pr)
+}
+# 최종 비교 그래프 
+price <- sapply(1:100, func) # 1~100반복 
+plot(price, type="l", xlab = "Number of steps", ylab = "Option Value")
+bs_price <- GBSOption(TypeFlag = "c", S = 900, X = 950, Time = 1/4, r = 0.02, sigma = 0.22, b = 0.02)@price
+abline(h = bs_price, col = 'red')
+legend("topright", legend = c('CRR-price', 'BS-price'), col = c('black', 'red'), pch = 19)
+title(main = "Call Option Pricing models")
+
+## Greeks
+GBSGreeks(Selection = "delta", TypeFlag = "c", S = 900, X = 950, Time = 1/4, r = 0.02, b = 0.02, sigma = 0.22)
+GBSGreeks(Selection = "gamma", TypeFlag = "c", S = 900, X = 950, Time = 1/4, r = 0.02, b = 0.02, sigma  =0.22)
+#
+portfolio <- sapply(c('c', 'p'), function(otype){sapply(500:1500, function(price){
+  GBSGreeks(Selection = 'delta',
+            TypeFlag = otype,
+            S=price, X = 950,
+            Time = 1/4, r = 0.02,
+            b = 0.02,
+            sigma =0.22)
+})
+})
+
+head(portfolio)
+
+## Straddle Delta
+plot(500:1500, rowSums(portfolio), type = 'l', xlab = 'underlying Price', ylab = 'Straddle Delta')
+
+##Implied Volatility ###
+
+iv <- EuropeanOptionImpliedVolatility("call", 11.10, 100, 100, 0.01, 0.03 ,05, 0.4)
+#순서 (옵션type, 옵션가격, 기초자산, 행사가격, 배당수익률, 무위험수익률, 잔존만기, 변동성initial)
+iv
+iv_a <- AmericanOptionImpliedVolatility("call", 11.10, 100, 100, 0.01, 0.03 ,05, 0.4)
+iv_a
+
+### 9-2 신용파생상품  ###
+rc <- c("AAA", "AA", "A", "BBB", "BB", "B", "CCC", "Default")
+M <- matrix(c(90.81, 8.33, 0.68, 0.06, 0.08, 0.02, 0.01, 0.01,
+              0.70, 90.65, 7.79, 0.64, 0.06, 0.13, 0.02, 0.01,
+              0.09, 2.27, 91.05, 5.52, 0.74, 0.26, 0.01, 0.06,
+              0.02, 0.33, 5.95, 85.93, 5.30, 1.17, 1.12, 0.18,
+              0.03, 0.14, 0.67, 7.73, 80.53, 8.84, 1.00, 1.06,
+              0.01, 0.11, 0.24, 0.43, 6.48, 83.46, 4.07, 5.20,
+              0.21, 0, 0.22, 1.30, 2.38, 11.24, 64.86, 19.79,
+              0, 0, 0, 0, 0, 0, 0, 100
+              )/100, 8, 8, dimnames = list(rc, rc), byrow = TRUE)
+lgd <- 0.2 # 부도시손실률
+cm.cs(M, lgd) #신용스프레드
+#
+ead <- c(140000,100000,100000) # 부도시 exposure
+N <- 3 # firms
+n <- 50000 #난수
+r <- 0.03 #rf
+rating <- c("BBB", "AA", "B")
+firmnames <- c("Blizzard", "Activision", "Nexon")
+alpha <- 0.99
+rho <- matrix(c(1,0.4,0.6,0.4,1,0.5,0.6,0.5,1),3,3,dimnames=list(firmnames, firmnames),byrow=TRUE)
+cm.CVaR(M, lgd, ead, N, n,r,rho,alpha,rating) #신용VaR
+pnl <- cm.gain(M, lgd, ead, N, n, r, rho,rating) # 시뮬레이션 신용VaR
+pnl
+
+## CDS ##
+
+yct = c(1,2,3,4,5,7) #테너 (테너 : 채무발생일 ~ 만기일간 기한, 결제기간이라고함) 
+ycr = c(0.0050, 0.0070, 0.0080, 0.0100, 0.0120, 0.0150) #수익률곡선할인률 
+cct = c(1,3,5,7) #테너 
+ccsp = c(0.99, 0.98, 0.95, 0.92) #생존확률 
+tenors = c(1,3,5,7) #만기
+r = 0.4 #회수율 
+priceCDS(yct, ycr, cct, ccsp, tenors, r)
+#
+cdsSpreads = c(0.0050, 0.0070, 0.0090, 0.0110)
+bootstrapCDS(yct,ycr,cct,ccsp,r)
+
+## 금리파생상품 ##
+irswapvalue()
+
+
+### 9-3 Exotic Option ###
+
+## 9-3-1 Asian Option  - 평균가 ##
+#1번 방법 
+price <- GeometricAverageRateOption("c", 110, 120, 0.5, 0.03, 0.05 ,0.1)
+price
+#2번 방법
+TurnbullWakemanAsianApproxOption(TypeFlag = "p", S = 100, SA = 102, X = 120, Time = 0.5,
+                                 time = 0.25, tau = 0, r = 0.03, b = 0.05, sigma = 0.1)@price
+#3번 방법
+LevyAsianApproxOption(TypeFlag = "p", S = 100, SA = 102, X = 120, Time = 0.5,
+                      time = 0.25, r = 0.03, b = 0.05, sigma = 0.1)@price
+
+## 9-3-2 Barrier Option ##
+
+## 1번 Down and Out
+StandardBarrierOption(TypeFlag = "cdo", S  =100, X=90, H = 95, K = 3, Time=0.5, r = 0.08, b = 0.04, sigma = 0.25)@price
+#H는 경계값, K는 리베이트값
+
+##2번 Up and out down and out call (Double Barrier)
+DoubleBarrierOption(TypeFlag = "co", S = 100,X=100, L = 50, U = 150, Time =0.25,
+                    r = 0.1 ,b=0.1, sigma = 0.15, delta1 = -0.1, delta2 = 0.1)@price
+#L하한, U상한, 그에 맞는 Delta1, Delta2
+
+##3번 룩백 Barrier Up and out
+LookBarrierOption(TypeFlag = "cuo", S = 100, X = 100, H = 130, time1 = 0.25, Time2 = 1, r=0.1,b=0.1, sigma=0.15)@price
+
+##4번 Gap Digital
+GapOption(TypeFlag = "c", S = 50, X1= 50, X2= 57, Time = 0.5, r = 0.09, b = 0.09, sigma = 0.20)
+#두개의 행사가격
+
+## 5번 Cash or Nothing : 만기시점에 기초자산가격이 행사가격에 도달할 경우 미리 정한가 지불 ##
+CashOrNothingOption(TypeFlag = "p", S = 100, X = 80, K = 10, Time = 9/12, r = 0.06, b = 0, sigma = 0.35)
+
+## 6번 Two ASSet cash or nothing down-up 옵션 ##
+# 1번 자산 가격 > 행사가격 & 2번 자산가격 > 행사가격 일 경우 고정된 현금 지불
+
+TwoAssetCashOrNothingOption(TypeFlag = "c", S1 = 100, S2 = 100, X1 = 110, X2 = 90,
+                            K = 10, Time = 0.5, r=0.1, b1 = 0.05, b2= 0.06, sigma1 = 0.2, sigma2=0.25,
+                            rho = 0.5)@price
+# K는 만기시 지불하는 그 고정현금액
+# rho는 상관관계 
\ No newline at end of file
diff --git a/Learning Quantitative Finance with R.Rproj b/Learning Quantitative Finance with R.Rproj
new file mode 100644
index 0000000..3af27f6
--- /dev/null
+++ b/Learning Quantitative Finance with R.Rproj	
@@ -0,0 +1,13 @@
+Version: 1.0
+
+RestoreWorkspace: Default
+SaveWorkspace: Default
+AlwaysSaveHistory: Default
+
+EnableCodeIndexing: Yes
+UseSpacesForTab: Yes
+NumSpacesForTab: 2
+Encoding: UTF-8
+
+RnwWeave: Sweave
+LaTeX: pdfLaTeX