Nothing
R/arigamyannsvr.R
In AriGaMyANNSVR: Hybrid ARIMA-GARCH and Two Specially Designed ML-Based Models
Defines functions
my_svr
my_ann
ariga
series_stationarity
series_nonlinearity
series_descstat
Documented in
ariga
my_ann
my_svr
series_descstat
series_nonlinearity
series_stationarity
#' @title Descriptive Statistics Of A Series
#' @description Provides descriptive statistics of a particular series. First column in the output result mentions 10 different statistics and second column contains the Statistics values of the particular series.
#' @param Y Univariate series
#' @import dplyr, psych, e1071, stats
#'
#' @return
#' \itemize{
#' \item desc_table - A table contains 10 descriptive statistics row-wise
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- series_descstat(Y)
#' @references
#' \itemize{
#'\item Garai, S., & Paul, R. K. (2023). Development of MCS based-ensemble models using CEEMDAN decomposition and machine intelligence. Intelligent Systems with Applications, 18, 200202.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
series_descstat <- function(Y){
sw_test <- shapiro.test(Y)
p_value <- sw_test$p.value
sw_stat <- round(sw_test$statistic, 3)
if (p_value <= 0.01) {
shapiro_wilk <- paste0("***")
} else if (p_value <= 0.05) {
shapiro_wilk <- paste0("**")
} else if (p_value <= 0.1) {
shapiro_wilk <- paste0("*")
} else {
shapiro_wilk <- paste0("")
}
# embedding
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
stats <- rbind(length(Y),
round(min(Y), 3),
round(max(Y), 3),
round(mean(Y), 3),
round(sd(Y), 3),
round(sd(y), 3),
round((sd(Y) / mean(Y) * 100), 3),
round((psych::skew(Y)), 3),
round(e1071::kurtosis(Y), 3),
paste0(sw_stat, shapiro_wilk))
desc_table <- data.frame(cbind(c('N', 'Minimum', 'Maximum', 'Mean', 'SD', 'Cond_SD', 'CV(%)',
'Skewness', 'Kurtosis', 'Shapiro-Wilk'), stats))
colnames(desc_table) <- c('Statistics', 'Values')
return(desc_table)
}
#'@title Non linearity test of a Data Frame
#' @description Performs non linearity test result for a series. Provides output as a single element (data frame) list. First column mentions different statistics (eps). Other columns are the Statistics values of the particular dimension.
#' @param df Data Frame with first column as serial number or date
#' @import DescribeDF
#'
#' @return
#' \itemize{
#' \item nonlinearity_list - A list with a single element (data frame) . Element is named as the name of the series provided. The element is such that first column mentions different statistics and other columns are the Statistics values of the particular dimension.
#' }
#' @export
#'
#' @examples
#' my_series <- rnorm(100, 100, 10)
#' nonlinearity <- series_nonlinearity(my_series)
#' nonlinearity$my_series
#' @references
#' \itemize{
#'\item Garai, S., & Paul, R. K. (2023). Development of MCS based-ensemble models using CEEMDAN decomposition and machine intelligence. Intelligent Systems with Applications, 18, 200202.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., ... & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
series_nonlinearity <- function(Y){
sl_no <- 1:length(Y)
df <- data.frame(sl_no = sl_no, X = Y)
colnames(df)[2] <- paste(deparse(substitute(Y)))
nonlinearity_list <- DescribeDF::df_nonlinearity(df)
return(nonlinearity_list)
}
#'@title Stationarity Tests Of A Series
#' @description Provides a list of three data frames: 'ADF', 'PP', 'KPSS'. Also indicates whether the data is stationary or not according to the null hypothesis of the corresponding tests.
#' @param Y Univariate time series
#' @import DescribeDF
#'
#' @return
#' \itemize{
#' \item stationarity_table - List of three data frames: 'ADF', 'PP', 'KPSS'
#' }
#' @export
#'
#' @examples
#' my_series <- rnorm(100, 100, 10)
#' series_stationarity(my_series)
#' @references
#' \itemize{
#'\item Garai, S., & Paul, R. K. (2023). Development of MCS based-ensemble models using CEEMDAN decomposition and machine intelligence. Intelligent Systems with Applications, 18, 200202.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., ... & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
series_stationarity <- function(Y){
sl_no <- 1:length(Y)
df <- data.frame(sl_no = sl_no, X = Y)
colnames(df)[2] <- paste(deparse(substitute(Y)))
stationarity_table <- DescribeDF::df_stationarity(df)
return(stationarity_table)
}
#' @title ARIMA-GARCH Hybrid Modeling
#' @description First fits the time series data by using ARIMA model. If the residuals are having "arch" effect, then GARCH is fitted. Based on the previously mentioned condition final prediction is obtained. More can be found from Paul and Garai (2021) <doi:10.1007/s00500-021-06087-4>.
#' @param Y Univariate time series
#' @param ratio Ratio of number of observations in training and testing sets
#' @param n_lag Lag of the provided time series data
#' @import DescribeDf
#' @return
#' \itemize{
#' \item Output_ariga: List of three data frames: predict_compare, forecast_compare, and metrics
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- ariga(Y, ratio = 0.8, n_lag = 4)
#' @references
#' \itemize{
#' \item Paul, R. K., & Garai, S. (2021). Performance comparison of wavelets-based machine learning technique for forecasting agricultural commodity prices. Soft Computing, 25(20), 12857-12873.
#' \item Paul, R. K., & Garai, S. (2022). Wavelets based artificial neural network technique for forecasting agricultural prices. Journal of the Indian Society for Probability and Statistics, 23(1), 47-61.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
ariga <- function(Y, ratio = 0.9, n_lag = 4){
# embedding for finding log return
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
# Compute the average of non-zero contents in the data
nonzero_data <- y[y != 0 & !is.na(y)]
average_nonzero <- mean(nonzero_data)
# Replace NaNs with the average of non-zero contents in the data
y[is.nan(y)] <- average_nonzero
# Check the result
y
# embedding for finding lag series of actual series
embed_size_y <- n_lag+1 # same lag (lags_y-1) for every sub series
diff.2 <- embed(Yt,embed_size_y)
dim(diff.2)
Y_actual <- diff.2[,1]
Y_actual_1 <- diff.2[,2]
# train-test split
n <- length(Y_actual) # this is already (original length-embed_y)
Y_train <- Y_actual[1:(n*ratio)]
Y_test <- Y_actual[(n*ratio+1):n]
Y_train_1 <- Y_actual_1[1:(n*ratio)]
Y_test_1 <- Y_actual_1[(n*ratio+1):n]
# embedding for finding lag series of log return series
diff.3 <- embed(y, embed_size_y)
y_actual <- diff.3[,1]
y_train <- y_actual[1:(n*ratio)]
y_test <- y_actual[(n*ratio+1):n]
# ARIMA
arima_ori <- forecast::auto.arima(y_train)
order_ori <- forecast::arimaorder(arima_ori)
model_arima_ori<-arima(y_train,order=c(order_ori[1], order_ori[2], order_ori[3]))
pred_arima_ori <-arima_ori$fitted
forecast_arima_ori <- data.frame(predict(arima_ori,n.ahead=(n-(n*ratio))))
forecast_arima_ori <-forecast_arima_ori$pred
#GARCH Model ####
ARCH_pvalue_ori <- as.numeric(FinTS::ArchTest(model_arima_ori$residuals)$p.value)
#ARCH_pvalue<-1
if(ARCH_pvalue_ori<=0.05){
garch.fit_ori <- fGarch::garchFit(~garch(1, 1), data = y_train, trace = FALSE)
pred_V_ori <- garch.fit_ori@fitted
forecast_V_ori <- predict(garch.fit_ori, n.ahead=(n-(n*ratio)))
forecast_V_ori <- forecast_V_ori$meanForecast
Resid_V_ori <- garch.fit_ori@residuals
for_resid_ori<-as.ts(y_test-forecast_V_ori)
}else {
pred_V_ori <- pred_arima_ori
forecast_V_ori <- forecast_arima_ori
Resid_V_ori <- as.ts(model_arima_ori$residuals)
for_resid_ori<-as.vector(y_test-as.vector(forecast_arima_ori))
}
ARIGA_train <- exp(pred_V_ori)*Y_train_1
ARIGA_test <- exp(forecast_V_ori)*Y_test_1
# validation_ARIGA
metrics_ariga_train <- data.frame(AllMetrics::all_metrics(Y_train, ARIGA_train))
metrics_ariga_test <- data.frame(AllMetrics::all_metrics(Y_test, ARIGA_test))
metrics <- cbind(metrics_ariga_train$Metrics,
as.numeric(metrics_ariga_train$Values),
as.numeric(metrics_ariga_test$Values))
colnames(metrics) <- c('Metrics', 'ARIGA_Train',
'ARIGA_Test')
predict_compare <- data.frame(cbind(train_actual = Y_train,
predicted = ARIGA_train))
colnames(predict_compare) <- c('train_actual', 'train_predicted')
forecast_compare <- data.frame(cbind(test_actual = Y_test,
forecast = ARIGA_test))
colnames(forecast_compare) <- c('test_actual', 'test_predicted')
Output_ariga <- list(predict_compare=predict_compare,
forecast_compare=forecast_compare,
metrics=metrics)
return(Output_ariga)
}
#' @title Specially Designed SVR-Based Modeling
#' @description Fits a ANN model to the uni-variate time series data. The contribution is related to the PhD work of the maintainer. More can be found from Paul and Garai (2021) <doi:10.1007/s00500-021-06087-4>.
#' @param Y Univariate time series
#' @param ratio Ratio of number of observations in training and testing sets
#' @param n_lag Lag of the provided time series data
#' @import forecast, AllMetrics, neuralnet
#' @return
#' \itemize{
#' \item Output_ann: List of three data frames: predict_compare, forecast_compare, and metrics
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- my_ann(Y, ratio = 0.8, n_lag = 4)
#' @references
#' \itemize{
#' \item Paul, R. K., & Garai, S. (2021). Performance comparison of wavelets-based machine learning technique for forecasting agricultural commodity prices. Soft Computing, 25(20), 12857-12873.
#' \item Paul, R. K., & Garai, S. (2022). Wavelets based artificial neural network technique for forecasting agricultural prices. Journal of the Indian Society for Probability and Statistics, 23(1), 47-61.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
my_ann <- function(Y, ratio = 0.9, n_lag = 4){
# embedding for finding log return
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
# Compute the average of non-zero contents in the data
nonzero_data <- y[y != 0 & !is.na(y)]
average_nonzero <- mean(nonzero_data)
# Replace NaNs with the average of non-zero contents in the data
y[is.nan(y)] <- average_nonzero
# Check the result
y
# embedding for finding lag series of actual series
embed_size_y <- n_lag+1 # same lag (lags_y-1) for every sub series
diff.2 <- embed(Yt,embed_size_y)
dim(diff.2)
Y_actual <- diff.2[,1]
Y_actual_1 <- diff.2[,2]
# train-test split
n <- length(Y_actual) # this is already (original length-embed_y)
Y_train <- Y_actual[1:(n*ratio)]
Y_test <- Y_actual[(n*ratio+1):n]
Y_train_1 <- Y_actual_1[1:(n*ratio)]
Y_test_1 <- Y_actual_1[(n*ratio+1):n]
# embedding for finding lag series of log return series
diff.3 <- embed(y, embed_size_y)
diff.3_train <- diff.3[1:(n*ratio),]
diff.3_test <- diff.3[(n*ratio+1):n,]
# Set the column names
colnames(diff.3_train) <- c("y_train", paste0("x", 1:(embed_size_y-1)))
colnames(diff.3_test) <- c("y_test", paste0("x", 1:(embed_size_y-1)))
# extract the predictors and response from the current matrix
predictors_train <- diff.3_train[, -1]
response_train <- diff.3_train[, 1]
allvars <- colnames(predictors_train)
predictorvars <- paste(allvars, collapse="+")
form <- as.formula(paste("y_train ~", predictorvars))
y_actual <- diff.3[,1]
y_train <- y_actual[1:(n*ratio)]
y_test <- y_actual[(n*ratio+1):n]
# ANN ####
nn <- neuralnet::neuralnet(form, diff.3_train, hidden = c(4))
train_predicted <- predict(nn, predictors_train)
predictors_test <- diff.3_test[, -1]
response_test <- diff.3_test[, 1]
test_predicted <- predict(nn, predictors_test)
ann_train <- exp(train_predicted)*Y_train_1
ann_test <- exp(test_predicted)*Y_test_1
# validation_ann
metrics_ann_train <- data.frame(AllMetrics:: all_metrics(Y_train, ann_train))
metrics_ann_test <- data.frame(AllMetrics:: all_metrics(Y_test, ann_test))
metrics <- cbind(metrics_ann_train$Metrics,
as.numeric(metrics_ann_train$Values),
as.numeric(metrics_ann_test$Values))
colnames(metrics) <- c('Metrics', 'ANN_Train',
'ANN_Test')
predict_compare <- data.frame(cbind(train_actual = Y_train,
predicted = ann_train))
colnames(predict_compare) <- c('train_actual', 'train_predicted')
forecast_compare <- data.frame(cbind(test_actual = Y_test,
forecast = ann_test))
colnames(forecast_compare) <- c('test_actual', 'test_predicted')
Output_ann <- list(predict_compare=predict_compare,
forecast_compare=forecast_compare,
metrics=metrics)
return(Output_ann)
}
#' @title Specially Designed SVR-Based Modeling
#' @description Fits a SVR model to the uni-variate time series data. The contribution is related to the PhD work of the maintainer. More can be found from Paul and Garai (2021) <doi:10.1007/s00500-021-06087-4>.
#' @param Y Univariate time series
#' @param ratio Ratio of number of observations in training and testing sets
#' @param n_lag Lag of the provided time series data
#' @import forecast, AllMetrics, e1071
#' @return
#' \itemize{
#' \item Output_ann: List of three data frames: predict_compare, forecast_compare, and metrics
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- my_svr(Y, ratio = 0.8, n_lag = 4)
#' @references
#' \itemize{
#' \item Paul, R. K., & Garai, S. (2021). Performance comparison of wavelets-based machine learning technique for forecasting agricultural commodity prices. Soft Computing, 25(20), 12857-12873.
#' \item Paul, R. K., & Garai, S. (2022). Wavelets based artificial neural network technique for forecasting agricultural prices. Journal of the Indian Society for Probability and Statistics, 23(1), 47-61.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
my_svr <- function(Y, ratio = 0.9, n_lag = 4){
# embedding for finding log return
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
# Compute the average of non-zero contents in the data
nonzero_data <- y[y != 0 & !is.na(y)]
average_nonzero <- mean(nonzero_data)
# Replace NaNs with the average of non-zero contents in the data
y[is.nan(y)] <- average_nonzero
# Check the result
y
# embedding for finding lag series of actual series
embed_size_y <- n_lag+1 # same lag (lags_y-1) for every sub series
diff.2 <- embed(Yt,embed_size_y)
dim(diff.2)
Y_actual <- diff.2[,1]
Y_actual_1 <- diff.2[,2]
# train-test split
n <- length(Y_actual) # this is already (original length-embed_y)
Y_train <- Y_actual[1:(n*ratio)]
Y_test <- Y_actual[(n*ratio+1):n]
Y_train_1 <- Y_actual_1[1:(n*ratio)]
Y_test_1 <- Y_actual_1[(n*ratio+1):n]
# embedding for finding lag series of log return series
diff.3 <- embed(y, embed_size_y)
diff.3_train <- diff.3[1:(n*ratio),]
diff.3_test <- diff.3[(n*ratio+1):n,]
# Set the column names
colnames(diff.3_train) <- c("y_train", paste0("x", 1:(embed_size_y-1)))
colnames(diff.3_test) <- c("y_test", paste0("x", 1:(embed_size_y-1)))
# extract the predictors and response from the current matrix
predictors_train <- diff.3_train[, -1]
response_train <- diff.3_train[, 1]
allvars <- colnames(predictors_train)
predictorvars <- paste(allvars, collapse="+")
form <- as.formula(paste("y_train ~", predictorvars))
y_actual <- diff.3[,1]
y_train <- y_actual[1:(n*ratio)]
y_test <- y_actual[(n*ratio+1):n]
# SVR ####
# train a svm model with radial basis function as kernel
svmmodel <- e1071::svm(form, diff.3_train, type="eps-regression", kernel = 'radial')
# predict the response for the current matrix in the training set
train_predicted <- as.numeric(predict(svmmodel, predictors_train))
predictors_test <- diff.3_test[, -1]
response_test <- diff.3_test[, 1]
# predict the response for the current matrix in the testing set
test_predicted <- as.numeric(predict(svmmodel, predictors_test))
svr_train <- exp(train_predicted)*Y_train_1
svr_test <- exp(test_predicted)*Y_test_1
# validation_svr
metrics_svr_train <- data.frame(AllMetrics:: all_metrics(Y_train, svr_train))
metrics_svr_test <- data.frame(AllMetrics:: all_metrics(Y_test, svr_test))
metrics <- cbind(metrics_svr_train$Metrics,
as.numeric(metrics_svr_train$Values),
as.numeric(metrics_svr_test$Values))
colnames(metrics) <- c('Metrics', 'SVR_Train',
'SVR_Test')
predict_compare <- data.frame(cbind(train_actual = Y_train,
predicted = svr_train))
colnames(predict_compare) <- c('train_actual', 'train_predicted')
forecast_compare <- data.frame(cbind(test_actual = Y_test,
forecast = svr_test))
colnames(forecast_compare) <- c('test_actual', 'test_predicted')
Output_svr <- list(predict_compare=predict_compare,
forecast_compare=forecast_compare,
metrics=metrics)
return(Output_svr)
}
Try the AriGaMyANNSVR package in your browser
Any scripts or data that you put into this service are public.
AriGaMyANNSVR documentation built on April 14, 2023, 12:36 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions my_svr my_ann ariga series_stationarity series_nonlinearity series_descstat
Documented in ariga my_ann my_svr series_descstat series_nonlinearity series_stationarity
#' @title Descriptive Statistics Of A Series
#' @description Provides descriptive statistics of a particular series. First column in the output result mentions 10 different statistics and second column contains the Statistics values of the particular series.
#' @param Y Univariate series
#' @import dplyr, psych, e1071, stats
#'
#' @return
#' \itemize{
#' \item desc_table - A table contains 10 descriptive statistics row-wise
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- series_descstat(Y)
#' @references
#' \itemize{
#'\item Garai, S., & Paul, R. K. (2023). Development of MCS based-ensemble models using CEEMDAN decomposition and machine intelligence. Intelligent Systems with Applications, 18, 200202.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
series_descstat <- function(Y){
sw_test <- shapiro.test(Y)
p_value <- sw_test$p.value
sw_stat <- round(sw_test$statistic, 3)
if (p_value <= 0.01) {
shapiro_wilk <- paste0("***")
} else if (p_value <= 0.05) {
shapiro_wilk <- paste0("**")
} else if (p_value <= 0.1) {
shapiro_wilk <- paste0("*")
} else {
shapiro_wilk <- paste0("")
}
# embedding
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
stats <- rbind(length(Y),
round(min(Y), 3),
round(max(Y), 3),
round(mean(Y), 3),
round(sd(Y), 3),
round(sd(y), 3),
round((sd(Y) / mean(Y) * 100), 3),
round((psych::skew(Y)), 3),
round(e1071::kurtosis(Y), 3),
paste0(sw_stat, shapiro_wilk))
desc_table <- data.frame(cbind(c('N', 'Minimum', 'Maximum', 'Mean', 'SD', 'Cond_SD', 'CV(%)',
'Skewness', 'Kurtosis', 'Shapiro-Wilk'), stats))
colnames(desc_table) <- c('Statistics', 'Values')
return(desc_table)
}
#'@title Non linearity test of a Data Frame
#' @description Performs non linearity test result for a series. Provides output as a single element (data frame) list. First column mentions different statistics (eps). Other columns are the Statistics values of the particular dimension.
#' @param df Data Frame with first column as serial number or date
#' @import DescribeDF
#'
#' @return
#' \itemize{
#' \item nonlinearity_list - A list with a single element (data frame) . Element is named as the name of the series provided. The element is such that first column mentions different statistics and other columns are the Statistics values of the particular dimension.
#' }
#' @export
#'
#' @examples
#' my_series <- rnorm(100, 100, 10)
#' nonlinearity <- series_nonlinearity(my_series)
#' nonlinearity$my_series
#' @references
#' \itemize{
#'\item Garai, S., & Paul, R. K. (2023). Development of MCS based-ensemble models using CEEMDAN decomposition and machine intelligence. Intelligent Systems with Applications, 18, 200202.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., ... & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
series_nonlinearity <- function(Y){
sl_no <- 1:length(Y)
df <- data.frame(sl_no = sl_no, X = Y)
colnames(df)[2] <- paste(deparse(substitute(Y)))
nonlinearity_list <- DescribeDF::df_nonlinearity(df)
return(nonlinearity_list)
}
#'@title Stationarity Tests Of A Series
#' @description Provides a list of three data frames: 'ADF', 'PP', 'KPSS'. Also indicates whether the data is stationary or not according to the null hypothesis of the corresponding tests.
#' @param Y Univariate time series
#' @import DescribeDF
#'
#' @return
#' \itemize{
#' \item stationarity_table - List of three data frames: 'ADF', 'PP', 'KPSS'
#' }
#' @export
#'
#' @examples
#' my_series <- rnorm(100, 100, 10)
#' series_stationarity(my_series)
#' @references
#' \itemize{
#'\item Garai, S., & Paul, R. K. (2023). Development of MCS based-ensemble models using CEEMDAN decomposition and machine intelligence. Intelligent Systems with Applications, 18, 200202.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., ... & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
series_stationarity <- function(Y){
sl_no <- 1:length(Y)
df <- data.frame(sl_no = sl_no, X = Y)
colnames(df)[2] <- paste(deparse(substitute(Y)))
stationarity_table <- DescribeDF::df_stationarity(df)
return(stationarity_table)
}
#' @title ARIMA-GARCH Hybrid Modeling
#' @description First fits the time series data by using ARIMA model. If the residuals are having "arch" effect, then GARCH is fitted. Based on the previously mentioned condition final prediction is obtained. More can be found from Paul and Garai (2021) <doi:10.1007/s00500-021-06087-4>.
#' @param Y Univariate time series
#' @param ratio Ratio of number of observations in training and testing sets
#' @param n_lag Lag of the provided time series data
#' @import DescribeDf
#' @return
#' \itemize{
#' \item Output_ariga: List of three data frames: predict_compare, forecast_compare, and metrics
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- ariga(Y, ratio = 0.8, n_lag = 4)
#' @references
#' \itemize{
#' \item Paul, R. K., & Garai, S. (2021). Performance comparison of wavelets-based machine learning technique for forecasting agricultural commodity prices. Soft Computing, 25(20), 12857-12873.
#' \item Paul, R. K., & Garai, S. (2022). Wavelets based artificial neural network technique for forecasting agricultural prices. Journal of the Indian Society for Probability and Statistics, 23(1), 47-61.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
ariga <- function(Y, ratio = 0.9, n_lag = 4){
# embedding for finding log return
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
# Compute the average of non-zero contents in the data
nonzero_data <- y[y != 0 & !is.na(y)]
average_nonzero <- mean(nonzero_data)
# Replace NaNs with the average of non-zero contents in the data
y[is.nan(y)] <- average_nonzero
# Check the result
y
# embedding for finding lag series of actual series
embed_size_y <- n_lag+1 # same lag (lags_y-1) for every sub series
diff.2 <- embed(Yt,embed_size_y)
dim(diff.2)
Y_actual <- diff.2[,1]
Y_actual_1 <- diff.2[,2]
# train-test split
n <- length(Y_actual) # this is already (original length-embed_y)
Y_train <- Y_actual[1:(n*ratio)]
Y_test <- Y_actual[(n*ratio+1):n]
Y_train_1 <- Y_actual_1[1:(n*ratio)]
Y_test_1 <- Y_actual_1[(n*ratio+1):n]
# embedding for finding lag series of log return series
diff.3 <- embed(y, embed_size_y)
y_actual <- diff.3[,1]
y_train <- y_actual[1:(n*ratio)]
y_test <- y_actual[(n*ratio+1):n]
# ARIMA
arima_ori <- forecast::auto.arima(y_train)
order_ori <- forecast::arimaorder(arima_ori)
model_arima_ori<-arima(y_train,order=c(order_ori[1], order_ori[2], order_ori[3]))
pred_arima_ori <-arima_ori$fitted
forecast_arima_ori <- data.frame(predict(arima_ori,n.ahead=(n-(n*ratio))))
forecast_arima_ori <-forecast_arima_ori$pred
#GARCH Model ####
ARCH_pvalue_ori <- as.numeric(FinTS::ArchTest(model_arima_ori$residuals)$p.value)
#ARCH_pvalue<-1
if(ARCH_pvalue_ori<=0.05){
garch.fit_ori <- fGarch::garchFit(~garch(1, 1), data = y_train, trace = FALSE)
pred_V_ori <- garch.fit_ori@fitted
forecast_V_ori <- predict(garch.fit_ori, n.ahead=(n-(n*ratio)))
forecast_V_ori <- forecast_V_ori$meanForecast
Resid_V_ori <- garch.fit_ori@residuals
for_resid_ori<-as.ts(y_test-forecast_V_ori)
}else {
pred_V_ori <- pred_arima_ori
forecast_V_ori <- forecast_arima_ori
Resid_V_ori <- as.ts(model_arima_ori$residuals)
for_resid_ori<-as.vector(y_test-as.vector(forecast_arima_ori))
}
ARIGA_train <- exp(pred_V_ori)*Y_train_1
ARIGA_test <- exp(forecast_V_ori)*Y_test_1
# validation_ARIGA
metrics_ariga_train <- data.frame(AllMetrics::all_metrics(Y_train, ARIGA_train))
metrics_ariga_test <- data.frame(AllMetrics::all_metrics(Y_test, ARIGA_test))
metrics <- cbind(metrics_ariga_train$Metrics,
as.numeric(metrics_ariga_train$Values),
as.numeric(metrics_ariga_test$Values))
colnames(metrics) <- c('Metrics', 'ARIGA_Train',
'ARIGA_Test')
predict_compare <- data.frame(cbind(train_actual = Y_train,
predicted = ARIGA_train))
colnames(predict_compare) <- c('train_actual', 'train_predicted')
forecast_compare <- data.frame(cbind(test_actual = Y_test,
forecast = ARIGA_test))
colnames(forecast_compare) <- c('test_actual', 'test_predicted')
Output_ariga <- list(predict_compare=predict_compare,
forecast_compare=forecast_compare,
metrics=metrics)
return(Output_ariga)
}
#' @title Specially Designed SVR-Based Modeling
#' @description Fits a ANN model to the uni-variate time series data. The contribution is related to the PhD work of the maintainer. More can be found from Paul and Garai (2021) <doi:10.1007/s00500-021-06087-4>.
#' @param Y Univariate time series
#' @param ratio Ratio of number of observations in training and testing sets
#' @param n_lag Lag of the provided time series data
#' @import forecast, AllMetrics, neuralnet
#' @return
#' \itemize{
#' \item Output_ann: List of three data frames: predict_compare, forecast_compare, and metrics
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- my_ann(Y, ratio = 0.8, n_lag = 4)
#' @references
#' \itemize{
#' \item Paul, R. K., & Garai, S. (2021). Performance comparison of wavelets-based machine learning technique for forecasting agricultural commodity prices. Soft Computing, 25(20), 12857-12873.
#' \item Paul, R. K., & Garai, S. (2022). Wavelets based artificial neural network technique for forecasting agricultural prices. Journal of the Indian Society for Probability and Statistics, 23(1), 47-61.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
my_ann <- function(Y, ratio = 0.9, n_lag = 4){
# embedding for finding log return
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
# Compute the average of non-zero contents in the data
nonzero_data <- y[y != 0 & !is.na(y)]
average_nonzero <- mean(nonzero_data)
# Replace NaNs with the average of non-zero contents in the data
y[is.nan(y)] <- average_nonzero
# Check the result
y
# embedding for finding lag series of actual series
embed_size_y <- n_lag+1 # same lag (lags_y-1) for every sub series
diff.2 <- embed(Yt,embed_size_y)
dim(diff.2)
Y_actual <- diff.2[,1]
Y_actual_1 <- diff.2[,2]
# train-test split
n <- length(Y_actual) # this is already (original length-embed_y)
Y_train <- Y_actual[1:(n*ratio)]
Y_test <- Y_actual[(n*ratio+1):n]
Y_train_1 <- Y_actual_1[1:(n*ratio)]
Y_test_1 <- Y_actual_1[(n*ratio+1):n]
# embedding for finding lag series of log return series
diff.3 <- embed(y, embed_size_y)
diff.3_train <- diff.3[1:(n*ratio),]
diff.3_test <- diff.3[(n*ratio+1):n,]
# Set the column names
colnames(diff.3_train) <- c("y_train", paste0("x", 1:(embed_size_y-1)))
colnames(diff.3_test) <- c("y_test", paste0("x", 1:(embed_size_y-1)))
# extract the predictors and response from the current matrix
predictors_train <- diff.3_train[, -1]
response_train <- diff.3_train[, 1]
allvars <- colnames(predictors_train)
predictorvars <- paste(allvars, collapse="+")
form <- as.formula(paste("y_train ~", predictorvars))
y_actual <- diff.3[,1]
y_train <- y_actual[1:(n*ratio)]
y_test <- y_actual[(n*ratio+1):n]
# ANN ####
nn <- neuralnet::neuralnet(form, diff.3_train, hidden = c(4))
train_predicted <- predict(nn, predictors_train)
predictors_test <- diff.3_test[, -1]
response_test <- diff.3_test[, 1]
test_predicted <- predict(nn, predictors_test)
ann_train <- exp(train_predicted)*Y_train_1
ann_test <- exp(test_predicted)*Y_test_1
# validation_ann
metrics_ann_train <- data.frame(AllMetrics:: all_metrics(Y_train, ann_train))
metrics_ann_test <- data.frame(AllMetrics:: all_metrics(Y_test, ann_test))
metrics <- cbind(metrics_ann_train$Metrics,
as.numeric(metrics_ann_train$Values),
as.numeric(metrics_ann_test$Values))
colnames(metrics) <- c('Metrics', 'ANN_Train',
'ANN_Test')
predict_compare <- data.frame(cbind(train_actual = Y_train,
predicted = ann_train))
colnames(predict_compare) <- c('train_actual', 'train_predicted')
forecast_compare <- data.frame(cbind(test_actual = Y_test,
forecast = ann_test))
colnames(forecast_compare) <- c('test_actual', 'test_predicted')
Output_ann <- list(predict_compare=predict_compare,
forecast_compare=forecast_compare,
metrics=metrics)
return(Output_ann)
}
#' @title Specially Designed SVR-Based Modeling
#' @description Fits a SVR model to the uni-variate time series data. The contribution is related to the PhD work of the maintainer. More can be found from Paul and Garai (2021) <doi:10.1007/s00500-021-06087-4>.
#' @param Y Univariate time series
#' @param ratio Ratio of number of observations in training and testing sets
#' @param n_lag Lag of the provided time series data
#' @import forecast, AllMetrics, e1071
#' @return
#' \itemize{
#' \item Output_ann: List of three data frames: predict_compare, forecast_compare, and metrics
#' }
#' @export
#'
#' @examples
#' Y <- rnorm(100, 100, 10)
#' result <- my_svr(Y, ratio = 0.8, n_lag = 4)
#' @references
#' \itemize{
#' \item Paul, R. K., & Garai, S. (2021). Performance comparison of wavelets-based machine learning technique for forecasting agricultural commodity prices. Soft Computing, 25(20), 12857-12873.
#' \item Paul, R. K., & Garai, S. (2022). Wavelets based artificial neural network technique for forecasting agricultural prices. Journal of the Indian Society for Probability and Statistics, 23(1), 47-61.
#' \item Garai, S., Paul, R. K., Rakshit, D., Yeasin, M., Paul, A. K., Roy, H. S., Barman, S. & Manjunatha, B. (2023). An MRA Based MLR Model for Forecasting Indian Annual Rainfall Using Large Scale Climate Indices. International Journal of Environment and Climate Change, 13(5), 137-150.
#' }
my_svr <- function(Y, ratio = 0.9, n_lag = 4){
# embedding for finding log return
diff.1 <- embed(Y, 2)
Yt <- diff.1[,1]
Yt_1 <- diff.1[,2]
y <- log(Yt/Yt_1)
# Compute the average of non-zero contents in the data
nonzero_data <- y[y != 0 & !is.na(y)]
average_nonzero <- mean(nonzero_data)
# Replace NaNs with the average of non-zero contents in the data
y[is.nan(y)] <- average_nonzero
# Check the result
y
# embedding for finding lag series of actual series
embed_size_y <- n_lag+1 # same lag (lags_y-1) for every sub series
diff.2 <- embed(Yt,embed_size_y)
dim(diff.2)
Y_actual <- diff.2[,1]
Y_actual_1 <- diff.2[,2]
# train-test split
n <- length(Y_actual) # this is already (original length-embed_y)
Y_train <- Y_actual[1:(n*ratio)]
Y_test <- Y_actual[(n*ratio+1):n]
Y_train_1 <- Y_actual_1[1:(n*ratio)]
Y_test_1 <- Y_actual_1[(n*ratio+1):n]
# embedding for finding lag series of log return series
diff.3 <- embed(y, embed_size_y)
diff.3_train <- diff.3[1:(n*ratio),]
diff.3_test <- diff.3[(n*ratio+1):n,]
# Set the column names
colnames(diff.3_train) <- c("y_train", paste0("x", 1:(embed_size_y-1)))
colnames(diff.3_test) <- c("y_test", paste0("x", 1:(embed_size_y-1)))
# extract the predictors and response from the current matrix
predictors_train <- diff.3_train[, -1]
response_train <- diff.3_train[, 1]
allvars <- colnames(predictors_train)
predictorvars <- paste(allvars, collapse="+")
form <- as.formula(paste("y_train ~", predictorvars))
y_actual <- diff.3[,1]
y_train <- y_actual[1:(n*ratio)]
y_test <- y_actual[(n*ratio+1):n]
# SVR ####
# train a svm model with radial basis function as kernel
svmmodel <- e1071::svm(form, diff.3_train, type="eps-regression", kernel = 'radial')
# predict the response for the current matrix in the training set
train_predicted <- as.numeric(predict(svmmodel, predictors_train))
predictors_test <- diff.3_test[, -1]
response_test <- diff.3_test[, 1]
# predict the response for the current matrix in the testing set
test_predicted <- as.numeric(predict(svmmodel, predictors_test))
svr_train <- exp(train_predicted)*Y_train_1
svr_test <- exp(test_predicted)*Y_test_1
# validation_svr
metrics_svr_train <- data.frame(AllMetrics:: all_metrics(Y_train, svr_train))
metrics_svr_test <- data.frame(AllMetrics:: all_metrics(Y_test, svr_test))
metrics <- cbind(metrics_svr_train$Metrics,
as.numeric(metrics_svr_train$Values),
as.numeric(metrics_svr_test$Values))
colnames(metrics) <- c('Metrics', 'SVR_Train',
'SVR_Test')
predict_compare <- data.frame(cbind(train_actual = Y_train,
predicted = svr_train))
colnames(predict_compare) <- c('train_actual', 'train_predicted')
forecast_compare <- data.frame(cbind(test_actual = Y_test,
forecast = svr_test))
colnames(forecast_compare) <- c('test_actual', 'test_predicted')
Output_svr <- list(predict_compare=predict_compare,
forecast_compare=forecast_compare,
metrics=metrics)
return(Output_svr)
}
Try the AriGaMyANNSVR package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.