Renamed anomaly detection function and added finetune_loss parameter.

NAMESPACE

@@ -5,8 +5,8 @@ export(.nixtla_data_prep)
 export(.validate_exogenous)
 export(date_conversion)
 export(infer_frequency)
-export(nixtla_client_anomaly_detection)
 export(nixtla_client_cross_validation)
+export(nixtla_client_detect_anomalies)
 export(nixtla_client_forecast)
 export(nixtla_client_historic)
 export(nixtla_client_plot)

R/nixtla_client_cross_validation.R

@@ -11,6 +11,7 @@
 #' @param n_windows Number of windows to evaluate.
 #' @param step_size Step size between each cross validation window. If NULL, it will equal the forecast horizon (h).
 #' @param finetune_steps Number of steps used to finetune TimeGPT in the new data.
+#' @param finetune_loss Loss function to use for finetuning. Options are: "default", "mae", "mse", "rmse", "mape", and "smape".
 #' @param clean_ex_first Clean exogenous signal before making the forecasts using TimeGPT.
 #' @param model Model to use, either "timegpt-1" or "timegpt-1-long-horizon". Use "timegpt-1-long-horizon" if you want to forecast more than one seasonal period given the frequency of the data.
 #'
@@ -24,7 +25,7 @@
 #'   fcst <- nixtlar::nixtla_client_cross_validation(df, h = 8, id_col = "unique_id", n_windows = 5)
 #' }
 #'
-nixtla_client_cross_validation <- function(df, h=8, freq=NULL, id_col=NULL, time_col="ds", target_col="y", X_df=NULL, level=NULL, n_windows=1, step_size=NULL, finetune_steps=0, clean_ex_first=TRUE, model="timegpt-1"){
+nixtla_client_cross_validation <- function(df, h=8, freq=NULL, id_col=NULL, time_col="ds", target_col="y", X_df=NULL, level=NULL, n_windows=1, step_size=NULL, finetune_steps=0, finetune_loss="default", clean_ex_first=TRUE, model="timegpt-1"){
 
   # Prepare data ----
   names(df)[which(names(df) == time_col)] <- "ds"
@@ -55,6 +56,7 @@ nixtla_client_cross_validation <- function(df, h=8, freq=NULL, id_col=NULL, time
     n_windows = n_windows,
     step_size = step_size,
     finetune_steps = finetune_steps,
+    finetune_loss = finetune_loss,
     clean_ex_first = clean_ex_first
   )
 

R/nixtla_client_anomaly_detection.R → R/nixtla_client_detect_anomalies.R

@@ -19,7 +19,7 @@
 #'   fcst <- nixtlar::nixtla_client_anomaly_detection(df, id_col="unique_id")
 #' }
 #'
-nixtla_client_anomaly_detection <- function(df, freq=NULL, id_col=NULL, time_col="ds", target_col="y", level=c(99), clean_ex_first=TRUE, model="timegpt-1"){
+nixtla_client_detect_anomalies <- function(df, freq=NULL, id_col=NULL, time_col="ds", target_col="y", level=c(99), clean_ex_first=TRUE, model="timegpt-1"){
 
   # Prepare data ----
   names(df)[which(names(df) == time_col)] <- "ds"

R/nixtla_client_forecast.R

@@ -9,6 +9,7 @@
 #' @param X_df A tsibble or a data frame with future exogenous variables.
 #' @param level The confidence levels (0-100) for the prediction intervals.
 #' @param finetune_steps Number of steps used to finetune TimeGPT in the new data.
+#' @param finetune_loss Loss function to use for finetuning. Options are: "default", "mae", "mse", "rmse", "mape", and "smape".
 #' @param clean_ex_first Clean exogenous signal before making the forecasts using TimeGPT.
 #' @param add_history Return fitted values of the model.
 #' @param model Model to use, either "timegpt-1" or "timegpt-1-long-horizon". Use "timegpt-1-long-horizon" if you want to forecast more than one seasonal period given the frequency of the data.
@@ -23,7 +24,7 @@
 #'   fcst <- nixtlar::nixtla_client_forecast(df, h=8, id_col="unique_id", level=c(80,95))
 #' }
 #'
-nixtla_client_forecast <- function(df, h=8, freq=NULL, id_col=NULL, time_col="ds", target_col="y", X_df=NULL, level=NULL, finetune_steps=0, clean_ex_first=TRUE, add_history=FALSE, model="timegpt-1"){
+nixtla_client_forecast <- function(df, h=8, freq=NULL, id_col=NULL, time_col="ds", target_col="y", X_df=NULL, level=NULL, finetune_steps=0, finetune_loss="default", clean_ex_first=TRUE, add_history=FALSE, model="timegpt-1"){
 
   # Prepare data ----
   names(df)[which(names(df) == time_col)] <- "ds"
@@ -49,6 +50,7 @@ nixtla_client_forecast <- function(df, h=8, freq=NULL, id_col=NULL, time_col="ds
     y = y,
     freq = freq,
     finetune_steps = finetune_steps,
+    finetune_loss = finetune_loss,
     clean_ex_first = clean_ex_first
     )
 

R/nixtla_client_historic.R

@@ -7,6 +7,7 @@
 #' @param target_col Column that contains the target variable.
 #' @param level The confidence levels (0-100) for the prediction intervals.
 #' @param finetune_steps Number of steps used to finetune TimeGPT in the new data.
+#' @param finetune_loss Loss function to use for finetuning. Options are: "default", "mae", "mse", "rmse", "mape", and "smape".
 #' @param clean_ex_first Clean exogenous signal before making the forecasts using TimeGPT.
 #'
 #' @return TimeGPT's forecast for the in-sample period.
@@ -19,7 +20,7 @@
 #'   fcst <- nixtlar::nixtla_client_historic(df, id_col="unique_id", level=c(80,95))
 #' }
 #'
-nixtla_client_historic <- function(df, freq=NULL, id_col=NULL, time_col="ds", target_col="y", level=NULL, finetune_steps=0, clean_ex_first=TRUE){
+nixtla_client_historic <- function(df, freq=NULL, id_col=NULL, time_col="ds", target_col="y", level=NULL, finetune_steps=0, finetune_loss="default", clean_ex_first=TRUE){
 
   # Prepare data ----
   names(df)[which(names(df) == time_col)] <- "ds"
@@ -43,6 +44,7 @@ nixtla_client_historic <- function(df, freq=NULL, id_col=NULL, time_col="ds", ta
     y = y,
     freq = freq,
     finetune_steps = finetune_steps,
+    finetune_loss = finetune_loss,
     clean_ex_first = clean_ex_first
   )
 

R/nixtla_client_plot.R

@@ -1,4 +1,4 @@
-#' Plot forecast and insample values.
+#' Plot the output of the following nixtla_client functions: forecast, historic, anomaly_detection, and cross_validation.
 #'
 #' @param df A tsibble or a data frame with time series data (insample values).
 #' @param fcst A tsibble or a data frame with the TimeGPT point forecast and the prediction intervals (if available).
@@ -10,7 +10,7 @@
 #' @param max_insample_length Max number of insample observations to be plotted.
 #' @param plot_anomalies Whether or not to plot anomalies.
 #'
-#' @return Plot with the forecast and insample values
+#' @return Corresponding plot.
 #' @export
 #'
 #' @examples

cran-comments.md

@@ -1,18 +1,13 @@
-## Resubmission 9 Feb 2024  
-This is a resubmission. In this version I have: 
 
+## Resumbmission
 
-## Resubmission 3 Feb 2024
-This is a resubmission. In this version I have:
+This is a resumbmission. In this version, I have: 
 
-* Fixed issue with .Renviron file in vignette to comply with CRAN's policies. 
-
-## Resubmission 1 Feb 2024
-This is a resubmission. In this version I have:
-
-* Fixed issue with tests that require a secret token to make an API call
+- Updated names of functions. 
+- Fixed issues to comply with CRAN's policies. 
 
 ## R CMD check results
+
 0 errors | 0 warnings | 1 note
 
 * This is a new release.