From 71d5f98fb2860b604c35c2553a037c8ab76121a5 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Wed, 23 Nov 2022 14:53:58 +0100
Subject: [PATCH 01/21] Add vitigeoss dict to .gitignore

---
 .gitignore | 1 +
 1 file changed, 1 insertion(+)

diff --git a/.gitignore b/.gitignore
index d95b20cb..ce85ad32 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1 +1,2 @@
 out-logs/
+conf/vitigeoss-vars-dict.yml
-- 
GitLab


From 1f80bdf61685b8eea370a9f953e284100eae1441 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Wed, 23 Nov 2022 17:15:19 +0100
Subject: [PATCH 02/21] First review of recipe checks (WIP)

---
 tools/check_recipe.R | 174 +++++++++++++++++++++++--------------------
 1 file changed, 93 insertions(+), 81 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index 25536335..a25c5f80 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -1,65 +1,70 @@
-check_recipe <- function(recipe, logger) {
+check_recipe <- function(recipe) {
 
   # recipe: yaml recipe already read it
   # output: errors or the total number of workflow (vars x regions) to compute
 
-  info(logger, paste("Checking recipe", recipe$filename))
+  info(recipe$Run$logger, paste("Checking recipe", recipe$name))
 
   # ---------------------------------------------------------------------
   # ANALYSIS CHECKS
   # ---------------------------------------------------------------------
 
-  TIME_SETTINGS = c('sdate','leadtimemin','leadtimemax','hcst_start','hcst_end')
-  PARAMS = c('Horizon','Time','Variables','Region','Regrid','Workflow','Datasets')
-  HORIZONS <- c('Subseasonal','Seasonal','Decadal')
+  TIME_SETTINGS = c('sdate', 'ftime_min', 'ftime_max', 'hcst_start', 'hcst_end')
+  PARAMS = c('Horizon', 'Time', 'Variables', 'Region', 'Regrid', 'Workflow',
+	     'Datasets')
+  HORIZONS <- c('subseasonal', 'seasonal', 'decadal')
 
-  # create output dirs:
-  if (!any(names(recipe) %in% "Analysis")) {
-    error(logger, "The recipe should contain an element called 'Analysis'.")
+  # Check basic elements in recipe:Analysis:
+  if (!("Analysis" %in% names(recipe))) {
+    error(recipe$Run$logger,
+	  "The recipe must contain an element called 'Analysis'.")
   }
 
   if (!all(PARAMS %in% names(recipe$Analysis))) {
     error(logger, 
-          paste("The element 'Analysis' in the recipe should contain these",
-         "elements:", paste(PARAMS, collapse = " ")))
+          paste("The element 'Analysis' in the recipe must contain these",
+          "elements:", paste(PARAMS, collapse = ", ")))
   }
 
-  if (!any(HORIZONS %in% recipe$Analysis$Horizon)) {
+  if (!any(HORIZONS %in% tolower(recipe$Analysis$Horizon))) {
     error(logger,
-         "The element 'Horizon' in the recipe should be one of the followings:",
-         paste(HORIZONS, collapse = " "))
+          "The element 'Horizon' in the recipe must be one of the following:",
+	  paste(HORIZONS, collapse = ", "))
   }
   # Check temporal settings and
   # count the number of verifications
   if (!all(TIME_SETTINGS %in% names(recipe$Analysis$Time))) {
-    error(logger, 
-       paste("The element 'Time' in the recipe should contain these elements:",
+    error(logger,
+       paste("The element 'Time' in the recipe must contain these elements:",
              paste(TIME_SETTINGS, collapse = " ")))
   }
-  if (is.null(recipe$Analysis$Time$sdate$fcst_year) ||
-    recipe$Analysis$Time$sdate$fcst_year == 'None') {
+  ## TODO: Is this needed?
+  if (is.null(recipe$Analysis$Time$fcst_year) || 
+      recipe$Analysis$Time$fcst_year == 'None') {
     stream <- "hindcast"
-    recipe$Analysis$Time$sdate$fcst_year <- 'YYYY'
+    # recipe$Analysis$Time$fcst_year <- 'YYYY'
   } else {
     stream <- "fcst"
   }
-  if (length(recipe$Analysis$Time$sdate$fcst_day) > 1 &&
-      tolower(recipe$Analysis$Horizon) != "subseasonal") {
+  ## TODO: To be implemented in the future
+  # if (length(recipe$Analysis$Time$sdate$fcst_day) > 1 &&
+  #     tolower(recipe$Analysis$Horizon) != "subseasonal") {
+  #   warn(logger,
+  #        paste("Only subseasonal verification allows multiple forecast days."),
+  #              "Element fcst_day in recipe set as 1.")
+  #   recipe$Analysis$Time$sdate$fcst_day <- '01'
+  # }
+  ## TODO: Delete, this parameter was deprecated
+  # if (is.null(recipe$Analysis$Time$sdate$fcst_sday)) {
+  #   error(logger,
+  #         paste("The element 'fcst_sday' in the recipe should be defined."))
+  # }
+  if (is.null(recipe$Analysis$Time$fcst_syear)) {
     warn(logger,
-         paste("Only subseasonal verification allows multiple forecast days."),
-               "Element fcst_day in recipe set as 1.")
-    recipe$Analysis$Time$sdate$fcst_day <- '01'
+         paste("The element 'fcst_year' is not defined in the recipe.
+	       No forecast year will be used."))
   }
-  if (is.null(recipe$Analysis$Time$sdate$fcst_sday)) {
-    error(logger,
-          paste("The element 'fcst_sday' in the recipe should be defined."))
-  }
-  if (is.null(recipe$Analysis$Time$sdate$fcst_syear)) {
-    error(logger,
-          paste("The element 'fcst_syear' in the recipe should be defined."))
-  }
-
-  
+  ## TODO: Move this inside 'if'? 
   fcst.sdate <- NULL
   for (syear in recipe$Analysis$Time$sdate$fcst_syear) {
       for (sday in recipe$Analysis$Time$sdate$fcst_sday) {
@@ -72,22 +77,23 @@ check_recipe <- function(recipe, logger) {
   # Regrid checks:
   if (length(recipe$Analysis$Regrid) != 2) {
     error(logger,
-          "The 'Regrid' element should specified the 'method' and 'type'.")
+          "The 'Regrid' element should specify the 'method' and 'type'.")
     stop("EXECUTION FAILED")    
   }
-# more checks
+  # TODO: Add Workflow checks?
   # ...
   # calculate number of workflows to create for each variable and 
   if (length(recipe$Analysis$Horizon) > 1) {
-    error(logger, "Only 1 Horizon can be specified in the recipe")
-    stop("EXECUTION FAILED")
-  }
-  nvar <- length(recipe$Analysis$Variables)
-  if (nvar > 2) {
-    error(logger,
-          "Only two type of Variables can be listed: ECVs and Indicators.")
+    error(logger, "Only one single Horizon can be specified in the recipe")
     stop("EXECUTION FAILED")
   }
+  ## TODO: Refine this
+  # nvar <- length(recipe$Analysis$Variables)
+  # if (nvar > 2) {
+  #   error(logger,
+  #         "Only two type of Variables can be listed: ECVs and Indicators.")
+  #   stop("EXECUTION FAILED")
+  # }
   # remove NULL or None Indicators or ECVs from the recipe:
   if (!is.null(recipe$Analysis$Variables$Indicators) &&
       !is.list(recipe$Analysis$Variables$Indicators)) {
@@ -100,41 +106,44 @@ check_recipe <- function(recipe, logger) {
            -which(names(recipe$Analysis$Variables) == 'ECVs')]
   }
   # Only one Calibration method allowed:
+  ## TODO: Make sure this condition is correctly written
   if ((is.logical(recipe$Analysis$Workflow$Calibration[[1]]) && 
-                 recipe$Analysis$Workflow$Calibration[[1]] == FALSE) ||
-                 recipe$Analysis$Workflow$Calibration[[1]] == 'None' ||
-                 is.null(recipe$Analysis$Workflow$Calibration[[1]])) {
+                  recipe$Analysis$Workflow$Calibration[[1]] == FALSE) ||
+                  recipe$Analysis$Workflow$Calibration[[1]] == 'None' ||
+                  is.null(recipe$Analysis$Workflow$Calibration[[1]])) {
     warn(logger, 
          "There is no Calibration method selected, raw data verification.")
-    recipe$Analysis$Workflow$Calibration[[1]] <- FALSE
+    recipe$Analysis$Workflow$Calibration[[1]] <- 'raw'
   } else {
     # remove multiple calibration methods
     if (is.null(names(recipe$Analysis$Workflow$Calibration))) {
       error(logger,
-          "The 'Calibration' element should specified at least the 'method'.")
+            "The 'Calibration' element 'method' must be specified.")
       stop("EXECUTION FAILED")
     }
   }  
 
   if ("Region" %in% names(recipe$Analysis)) {
-    nregions <- length(recipe$Analysis$Region$Regional)
     limits <- c('latmin', 'latmax', 'lonmin', 'lonmax')
-    for (i in 1:length(recipe$Analysis$Region)) {
-      if (!all(limits %in% names(recipe$Analysis$Region[[i]]))) {
-        limits <- paste(limits, collapse = " ")
-        error(logger,
-           paste("Each region defined in element 'Regional'",
-                 "should have 4 elements:",
-                 limits))
-        stop("EXECUTION FAILED")
-      }
-      # are numeric? class list mode list 
+    if (!all(limits %in% names(recipe$Analysis$Region))) {
+      error(recipe$Run$logger, paste("There must be 4 elements in 'Region':",
+				     limits))
     }
+    ## TODO: Implement multiple regions
+    # nregions <- length(recipe$Analysis$Region$Regional)
+    # for (i in 1:length(recipe$Analysis$Region)) {
+    #   if (!all(limits %in% names(recipe$Analysis$Region[[i]]))) {
+    #     limits <- paste(limits, collapse = " ")
+    #     error(logger,
+    #        paste("Each region defined in element 'Regional'",
+    #              "should have 4 elements:",
+    #              limits))
+    #     stop("EXECUTION FAILED")
+    #   }
+    #   # are numeric? class list mode list 
+    # }
   } else {
-    error(logger,
-           paste("'Region'",
-                 "should be defined",
-                 limits))
+    error(recipe$Run$logger, "'Region' must be defined.")
         stop("EXECUTION FAILED")
   }
   
@@ -142,39 +151,42 @@ check_recipe <- function(recipe, logger) {
   # RUN CHECKS
   # ---------------------------------------------------------------------
 
-  RUN_FIELDS = c("Loglevel","Terminal","output_dir","code_dir")
-  LOG_LEVELS = c("INFO","DEBUG","WARNING","ERROR")
+  RUN_FIELDS = c("Loglevel", "Terminal", "output_dir", "code_dir")
+  LOG_LEVELS = c("INFO", "DEBUG", "WARN", "ERROR", "FATAL")
 
-  if (!any(names(recipe) %in% "Run")) {
-    error(logger, "The recipe should contain an element called 'Run'.")
+  if (!("Run" %in% names(recipe))) {
+    stop("The recipe must contain an element named 'Run'.")
   }
   if (!all(RUN_FIELDS %in% names(recipe$Run))) {
-    error(logger, paste0("Run should contain the fields: ", 
-                         paste(RUN_FIELDS,collapse=", "), "."))
+    error(recipe$Run$logger, paste("Recipe element 'Run' must contain",
+				   "all of the following fields:",
+                                   paste(RUN_FIELDS, collapse=", "), "."))
   }
   if (!is.character(recipe$Run$output_dir)) {
-    error(logger,
-          paste("The Run element 'output_dir' in", recipe$filename,"file ",
+    error(recipe$Run$logger,
+          paste("The Run element 'output_dir' in", recipe$name, "file ",
                 "should be a character string indicating the path ",
-                "where to save the outputs."))
+                "where the outputs should be saved."))
   }
   if (!is.character(recipe$Run$code_dir)) {
-    error(logger,
-          paste("The Run element 'code_dir' in", recipe$filename,"file ",
+    error(recipe$Run$logger,
+          paste("The Run element 'code_dir' in", recipe$name, "file ",
                 "should be a character string indicating the path ",
                 "where the code is."))
   }
   if (!is.logical(recipe$Run$Terminal)) {
-    error(logger,
-          paste("The Run element 'Terminal' in", recipe$filename,"file ",
-                "should be a boolean value indicating wether to print or not the log",
-                "in the terminal."))
+    error(recipe$Run$logger,
+          paste("The Run element 'Terminal' in", recipe$filename, "file ",
+                "should be a boolean value indicating whether or not to",
+                "print the logs in the terminal."))
   }
-  if (!is.character(recipe$Run$Loglevel) || !any(recipe$Run$Loglevel %in% LOG_LEVELS)) {
+  ## TODO: Review this case, since default value is allowed
+  if (!is.character(recipe$Run$Loglevel) || 
+      !any(recipe$Run$Loglevel %in% LOG_LEVELS)) {
     error(logger,
-          paste("The Run element 'Loglevel' in", recipe$filename,"file ",
-                "should be a character string indicating one of the levels available: ",
-                paste0(LOG_LEVELS,collapse='/')))
+          paste("The Run element 'Loglevel' in", recipe$name, "file ",
+                "should be a character string specifying one of the levels available: ",
+                paste0(LOG_LEVELS, collapse='/')))
   }
 
   # ---------------------------------------------------------------------
-- 
GitLab


From 3680e1cbca0c88df9d880a8e55a9b3fe69bb72e2 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Fri, 25 Nov 2022 10:20:18 +0100
Subject: [PATCH 03/21] Update check_recipe()

---
 .../testing_recipes/wrong_recipe_example.yml  |  44 +++++++
 tools/check_recipe.R                          | 124 +++++++++++-------
 2 files changed, 117 insertions(+), 51 deletions(-)
 create mode 100644 modules/Loading/testing_recipes/wrong_recipe_example.yml

diff --git a/modules/Loading/testing_recipes/wrong_recipe_example.yml b/modules/Loading/testing_recipes/wrong_recipe_example.yml
new file mode 100644
index 00000000..e5d59ad8
--- /dev/null
+++ b/modules/Loading/testing_recipes/wrong_recipe_example.yml
@@ -0,0 +1,44 @@
+Description:
+  Author: V. Agudetse
+  Info: Incomplete recipe with incorrect fields to test the recipe checker.
+
+Analysis:
+  Horizon: Seasoning
+  Variables:
+    name: tas
+    freq: monthly_mean
+  Petaflops:
+    System:
+      name: system7c3s
+    Multimodel: False
+    Reference:
+      name: era5
+  Time:
+    sdate: '1101'
+    fcst_year: '2020'
+    hcst_start: '1993'
+    hcst_end: '2016'
+    ftime_max: 6
+  Region:
+    latmax: 10
+    lonmin: 0
+    lonmax: 20
+  Regrid:
+    method: bilinear
+    type: to_system
+  Workflow:
+    Calibration:
+      method:
+    Skill:
+      metric: RPS RPSS
+    Probabilities:
+      percentiles: [[1/3, 2/3], [1/10, 9/10], [1/4, 2/4, 3/4]]
+    Indicators:
+      index: no
+  ncores: 7
+  remove_NAs: yes
+  Output_format: S2S4E
+Run:
+  Loglevel: INFO
+  Terminal: yes
+  output_dir: /esarchive/scratch/vagudets/repos/auto-s2s/out-logs/
diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index a25c5f80..dbbfd665 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -3,7 +3,7 @@ check_recipe <- function(recipe) {
   # recipe: yaml recipe already read it
   # output: errors or the total number of workflow (vars x regions) to compute
 
-  info(recipe$Run$logger, paste("Checking recipe", recipe$name))
+  info(recipe$Run$logger, paste("Checking recipe:", recipe$name))
 
   # ---------------------------------------------------------------------
   # ANALYSIS CHECKS
@@ -13,30 +13,36 @@ check_recipe <- function(recipe) {
   PARAMS = c('Horizon', 'Time', 'Variables', 'Region', 'Regrid', 'Workflow',
 	     'Datasets')
   HORIZONS <- c('subseasonal', 'seasonal', 'decadal')
+  # Define error status variable
+  error_status <- F
 
   # Check basic elements in recipe:Analysis:
   if (!("Analysis" %in% names(recipe))) {
     error(recipe$Run$logger,
 	  "The recipe must contain an element called 'Analysis'.")
+    error_status <- T
   }
 
   if (!all(PARAMS %in% names(recipe$Analysis))) {
-    error(logger, 
-          paste("The element 'Analysis' in the recipe must contain these",
-          "elements:", paste(PARAMS, collapse = ", ")))
+    error(recipe$Run$logger, 
+          paste0("The element 'Analysis' in the recipe must contain all of ",
+          "the following: ", paste(PARAMS, collapse = ", "), "."))
+    error_status <- T
   }
 
   if (!any(HORIZONS %in% tolower(recipe$Analysis$Horizon))) {
-    error(logger,
-          "The element 'Horizon' in the recipe must be one of the following:",
-	  paste(HORIZONS, collapse = ", "))
+    error(recipe$Run$logger,
+          paste0("The element 'Horizon' in the recipe must be one of the ",
+		"following: ", paste(HORIZONS, collapse = ", "), "."))
+    error_status <- T
   }
   # Check temporal settings and
   # count the number of verifications
   if (!all(TIME_SETTINGS %in% names(recipe$Analysis$Time))) {
-    error(logger,
-       paste("The element 'Time' in the recipe must contain these elements:",
-             paste(TIME_SETTINGS, collapse = " ")))
+    error(recipe$Run$logger,
+	  paste0("The element 'Time' in the recipe must contain all of the ",
+		 "following: ", paste(TIME_SETTINGS, collapse = ", "), "."))
+    error_status <- T
   }
   ## TODO: Is this needed?
   if (is.null(recipe$Analysis$Time$fcst_year) || 
@@ -49,48 +55,49 @@ check_recipe <- function(recipe) {
   ## TODO: To be implemented in the future
   # if (length(recipe$Analysis$Time$sdate$fcst_day) > 1 &&
   #     tolower(recipe$Analysis$Horizon) != "subseasonal") {
-  #   warn(logger,
+  #   warn(recipe$Run$logger,
   #        paste("Only subseasonal verification allows multiple forecast days."),
   #              "Element fcst_day in recipe set as 1.")
   #   recipe$Analysis$Time$sdate$fcst_day <- '01'
   # }
   ## TODO: Delete, this parameter was deprecated
   # if (is.null(recipe$Analysis$Time$sdate$fcst_sday)) {
-  #   error(logger,
+  #   error(recipe$Run$logger,
   #         paste("The element 'fcst_sday' in the recipe should be defined."))
   # }
-  if (is.null(recipe$Analysis$Time$fcst_syear)) {
-    warn(logger,
-         paste("The element 'fcst_year' is not defined in the recipe.
-	       No forecast year will be used."))
+  if (is.null(recipe$Analysis$Time$fcst_year)) {
+    warn(recipe$Run$logger,
+         paste("The element 'fcst_year' is not defined in the recipe.",
+	       "No forecast year will be used."))
   }
-  ## TODO: Move this inside 'if'? 
+  ## TODO: Adapt and move this inside 'if'? 
   fcst.sdate <- NULL
-  for (syear in recipe$Analysis$Time$sdate$fcst_syear) {
-      for (sday in recipe$Analysis$Time$sdate$fcst_sday) {
-        fcst.sdate <- c(fcst.sdate, 
-                             paste0(syear,
-                               sprintf("%04d", as.numeric(sday))))
-      }
+  for (syear in recipe$Analysis$Time$fcst_year) {
+    for (sday in recipe$Analysis$Time$sdate) {
+      fcst.sdate <- c(fcst.sdate, 
+                      paste0(syear,
+                             sprintf("%04d", as.numeric(sday))))
+    }
   }
   fcst.sdate <- list(stream = stream, fcst.sdate = fcst.sdate)   
   # Regrid checks:
   if (length(recipe$Analysis$Regrid) != 2) {
-    error(logger,
-          "The 'Regrid' element should specify the 'method' and 'type'.")
-    stop("EXECUTION FAILED")    
+    error(recipe$Run$logger,
+          "The 'Regrid' element must specify the 'method' and 'type'.")
+    error_status <- T
   }
   # TODO: Add Workflow checks?
   # ...
   # calculate number of workflows to create for each variable and 
   if (length(recipe$Analysis$Horizon) > 1) {
-    error(logger, "Only one single Horizon can be specified in the recipe")
-    stop("EXECUTION FAILED")
+    error(recipe$Run$logger,
+	  "Only one single Horizon can be specified in the recipe")
+    error_status <- T
   }
   ## TODO: Refine this
   # nvar <- length(recipe$Analysis$Variables)
   # if (nvar > 2) {
-  #   error(logger,
+  #   error(recipe$Run$logger,
   #         "Only two type of Variables can be listed: ECVs and Indicators.")
   #   stop("EXECUTION FAILED")
   # }
@@ -107,34 +114,36 @@ check_recipe <- function(recipe) {
   }
   # Only one Calibration method allowed:
   ## TODO: Make sure this condition is correctly written
-  if ((is.logical(recipe$Analysis$Workflow$Calibration[[1]]) && 
-                  recipe$Analysis$Workflow$Calibration[[1]] == FALSE) ||
-                  recipe$Analysis$Workflow$Calibration[[1]] == 'None' ||
-                  is.null(recipe$Analysis$Workflow$Calibration[[1]])) {
-    warn(logger, 
-         "There is no Calibration method selected, raw data verification.")
-    recipe$Analysis$Workflow$Calibration[[1]] <- 'raw'
+  if ((is.logical(recipe$Analysis$Workflow$Calibration$method) && 
+                  recipe$Analysis$Workflow$Calibration$method == FALSE) ||
+                  recipe$Analysis$Workflow$Calibration$method == 'None' ||
+                  is.null(recipe$Analysis$Workflow$Calibration$method)) {
+    warn(recipe$Run$logger,
+         "No Calibration method was specified, raw data verification.")
+    recipe$Analysis$Workflow$Calibration$method <- 'raw'
   } else {
     # remove multiple calibration methods
     if (is.null(names(recipe$Analysis$Workflow$Calibration))) {
-      error(logger,
+      error(recipe$Run$logger,
             "The 'Calibration' element 'method' must be specified.")
-      stop("EXECUTION FAILED")
+      error_status <- T
     }
-  }  
+  }
 
   if ("Region" %in% names(recipe$Analysis)) {
     limits <- c('latmin', 'latmax', 'lonmin', 'lonmax')
     if (!all(limits %in% names(recipe$Analysis$Region))) {
-      error(recipe$Run$logger, paste("There must be 4 elements in 'Region':",
-				     limits))
+      error(recipe$Run$logger,
+	    paste0("There must be 4 elements in 'Region': ",
+		   paste(limits, collapse = ", "), "."))
+      error_status <- T
     }
     ## TODO: Implement multiple regions
     # nregions <- length(recipe$Analysis$Region$Regional)
     # for (i in 1:length(recipe$Analysis$Region)) {
     #   if (!all(limits %in% names(recipe$Analysis$Region[[i]]))) {
     #     limits <- paste(limits, collapse = " ")
-    #     error(logger,
+    #     error(recipe$Run$logger,
     #        paste("Each region defined in element 'Regional'",
     #              "should have 4 elements:",
     #              limits))
@@ -144,7 +153,7 @@ check_recipe <- function(recipe) {
     # }
   } else {
     error(recipe$Run$logger, "'Region' must be defined.")
-        stop("EXECUTION FAILED")
+    error_status <- T
   }
   
   # ---------------------------------------------------------------------
@@ -161,24 +170,28 @@ check_recipe <- function(recipe) {
     error(recipe$Run$logger, paste("Recipe element 'Run' must contain",
 				   "all of the following fields:",
                                    paste(RUN_FIELDS, collapse=", "), "."))
+    error_status <- T
   }
   if (!is.character(recipe$Run$output_dir)) {
     error(recipe$Run$logger,
-          paste("The Run element 'output_dir' in", recipe$name, "file ",
-                "should be a character string indicating the path ",
-                "where the outputs should be saved."))
+          paste("The Run element 'output_dir' in", recipe$name, "file",
+                "should be a character string indicating the path where",
+                "the outputs should be saved."))
+    error_status <- T
   }
   if (!is.character(recipe$Run$code_dir)) {
     error(recipe$Run$logger,
           paste("The Run element 'code_dir' in", recipe$name, "file ",
                 "should be a character string indicating the path ",
                 "where the code is."))
+    error_status <- T
   }
   if (!is.logical(recipe$Run$Terminal)) {
     error(recipe$Run$logger,
-          paste("The Run element 'Terminal' in", recipe$filename, "file ",
+          paste("The Run element 'Terminal' in", recipe$name, "file ",
                 "should be a boolean value indicating whether or not to",
                 "print the logs in the terminal."))
+    error_status <- T
   }
   ## TODO: Review this case, since default value is allowed
   if (!is.character(recipe$Run$Loglevel) || 
@@ -187,6 +200,7 @@ check_recipe <- function(recipe) {
           paste("The Run element 'Loglevel' in", recipe$name, "file ",
                 "should be a character string specifying one of the levels available: ",
                 paste0(LOG_LEVELS, collapse='/')))
+    error_status <- T
   }
 
   # ---------------------------------------------------------------------
@@ -194,10 +208,18 @@ check_recipe <- function(recipe) {
   # ---------------------------------------------------------------------
   # Check workflow: need to define restrictions? 
   # e.g. only one calibration method
-  nverifications <- check_number_of_dependent_verifications(recipe)
-  info(logger, paste("Start Dates", paste(fcst.sdate, collapse = " ")))
-  info(logger, "Recipe checked succsessfully.")
-  return(append(nverifications, fcst.sdate))
+  ## TODO: Implement numbr of dependent verifications
+  #nverifications <- check_number_of_dependent_verifications(recipe)
+  info(recipe$Run$logger, paste("Start Dates:",
+				paste(fcst.sdate, collapse = " ")))
+
+  # Return error if any check has failed
+  if (error_status) {
+    stop("Recipe check failed. Please check the logs for a list of errors.")
+  } else {
+    info(recipe$Run$logger, "##### RECIPE CHECKS WERE SUCCESSFULL #####.")
+    # return(append(nverifications, fcst.sdate))
+  }
 }
 
 check_number_of_dependent_verifications <- function(recipe) {
-- 
GitLab


From cae758b0e71e818fc27f5b76acda85f5f01accc1 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Mon, 12 Dec 2022 15:25:19 +0100
Subject: [PATCH 04/21] Add example of main recipe

---
 recipes/recipe_example.yml | 58 ++++++++++++++++++++++++++++++++++++++
 tools/divide_recipe.R      |  6 ++--
 2 files changed, 62 insertions(+), 2 deletions(-)
 create mode 100644 recipes/recipe_example.yml

diff --git a/recipes/recipe_example.yml b/recipes/recipe_example.yml
new file mode 100644
index 00000000..27f74df9
--- /dev/null
+++ b/recipes/recipe_example.yml
@@ -0,0 +1,58 @@
+################################################################################ 
+## RECIPE DESCRIPTION  
+################################################################################ 
+
+Description:
+  Author: V. Agudetse
+  Info: Test for main recipe
+
+################################################################################
+## ANALYSIS CONFIGURATION
+################################################################################
+
+Analysis:
+  Horizon: Seasonal
+  Variables: # ECVs and Indicators?
+    - {name: tas, freq: monthly_mean}
+    - {name: prlr, freq: monthly_mean}
+  Datasets:
+    System: # multiple systems for single model, split if Multimodel = F
+      - {name: system7c3s}
+      - {name: system5c3s}
+    Multimodel: False # single option
+    Reference:
+      name: era5 # multiple references for single model?
+  Time:
+    sdate: '1101' # list, split
+    fcst_year: '2020' # list, don't split, handled internally
+    hcst_start: '1993' # single option
+    hcst_end: '2016' # single option
+    ftime_min: 1 # single option
+    ftime_max: 6 # singl eoption
+  Region: # multiple lists, split? Add region name if length(Region) > 1
+    - {name: "global", latmin: -90, latmax: 90, lonmin: 0, lonmax: 360}
+    - {name: "nino34", latmin: -5, latmax: 5, lonmin: -10, lonmax: 60}
+  Regrid:
+    method: bilinear # allow multiple methods?
+    type: to_system
+  Workflow:
+    Calibration:
+      method: mse_min # list, split?
+    Skill:
+      metric: RPS, RPSS, CRPS, CRPSS, FRPSS, BSS10, BSS90, mean_bias, mean_bias_SS # list, don't split
+    Probabilities:
+      percentiles: [[1/3, 2/3], [1/10, 9/10], [1/4, 2/4, 3/4]] # list, don't split
+    Indicators:
+      index: no # ? 
+  ncores: 7
+  remove_NAs: yes # bool, don't split
+  Output_format: S2S4E # string, don't split
+
+################################################################################
+## Run CONFIGURATION
+################################################################################
+Run:
+  Loglevel: INFO
+  Terminal: yes
+  output_dir: /esarchive/scratch/vagudets/repos/auto-s2s/out-logs/
+  code_dir: /esarchive/scratch/vagudets/repos/auto-s2s/
diff --git a/tools/divide_recipe.R b/tools/divide_recipe.R
index dafc8704..8f4ef35a 100644
--- a/tools/divide_recipe.R
+++ b/tools/divide_recipe.R
@@ -2,6 +2,8 @@
 # verifications: the output from check_recipe
 # folder: the name of the output folder for this run
 # logger: the log file obtain from prepare_outputs
+
+## TODO: Rethink strategy and adapt function accordingly
 divide_recipe <- function(recipe, verifications, folder, logger) {
   info(logger, "Spliting recipe in single verifications.")
   beta_recipe <- list(Description = append(recipe$Description, 
@@ -14,8 +16,8 @@ divide_recipe <- function(recipe, verifications, folder, logger) {
                                       Regrid = recipe$Analysis$Regrid,
                                       Workflow = recipe$Analysis$Workflow,
                                       Output_format = 
-                                              recipe$Analysis$Output_format),
-                       Run = recipe$Run)
+				        recipe$Analysis$Output_format),
+		      Run = recipe$Run)
   # duplicate recipe by Variables considering dep and indep:
   all_recipes <- list(beta_recipe)
   i <- 1 # to get track of the recipe number
-- 
GitLab


From fbb329af5930396261b8f8e4e28ff3cca2ddd106 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 13 Dec 2022 11:43:21 +0100
Subject: [PATCH 05/21] Formatting

---
 tools/check_recipe.R |  2 +-
 tools/data_summary.R | 11 +++++------
 2 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index dbbfd665..316b26f7 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -26,7 +26,7 @@ check_recipe <- function(recipe) {
   if (!all(PARAMS %in% names(recipe$Analysis))) {
     error(recipe$Run$logger, 
           paste0("The element 'Analysis' in the recipe must contain all of ",
-          "the following: ", paste(PARAMS, collapse = ", "), "."))
+		 "the following: ", paste(PARAMS, collapse = ", "), "."))
     error_status <- T
   }
 
diff --git a/tools/data_summary.R b/tools/data_summary.R
index 34b6bd6e..cddd4f9e 100644
--- a/tools/data_summary.R
+++ b/tools/data_summary.R
@@ -8,15 +8,15 @@ data_summary <- function(data_cube, recipe) {
   # Get name, leadtime months and date range
   object_name <- deparse(substitute(data_cube))
   if (recipe$Analysis$Variables$freq == "monthly_mean") {
-    date_format <- '%b %Y'
+    date_format <- "%b %Y"
   } else if (recipe$Analysis$Variables$freq == "daily_mean") {
-    date_format <- '%b %d %Y'
+    date_format <- "%b %d %Y"
   }
-  months <- unique(format(as.Date(data_cube$Dates[[1]]), format = '%B'))
-  months <- paste(as.character(months), collapse=", ")
+  months <- unique(format(as.Date(data_cube$Dates[[1]]), format = "%B"))
+  months <- paste(as.character(months), collapse = ", ")
   sdate_min <- format(min(as.Date(data_cube$Dates[[1]])), format = date_format)
   sdate_max <- format(max(as.Date(data_cube$Dates[[1]])), format = date_format)
-  
+
   # Create log instance and sink output to logfile and terminal
   info(recipe$Run$logger, "DATA SUMMARY:")
   sink(recipe$Run$logfile, append = TRUE, split = TRUE)
@@ -29,4 +29,3 @@ data_summary <- function(data_cube, recipe) {
   print("---------------------------------------------")
   sink()
 }
-  
-- 
GitLab


From 8da9afb6c3de5ba472b7aaf22f69c70a4d7662d3 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Thu, 15 Dec 2022 15:09:52 +0100
Subject: [PATCH 06/21] Update wrong example and recipe checker

---
 .../testing_recipes/wrong_recipe_example.yml  |  2 +-
 tools/check_recipe.R                          | 28 +++++++++++--------
 2 files changed, 18 insertions(+), 12 deletions(-)

diff --git a/modules/Loading/testing_recipes/wrong_recipe_example.yml b/modules/Loading/testing_recipes/wrong_recipe_example.yml
index e5d59ad8..12e2fc06 100644
--- a/modules/Loading/testing_recipes/wrong_recipe_example.yml
+++ b/modules/Loading/testing_recipes/wrong_recipe_example.yml
@@ -15,7 +15,7 @@ Analysis:
       name: era5
   Time:
     sdate: '1101'
-    fcst_year: '2020'
+    fcst_syear: '2020'
     hcst_start: '1993'
     hcst_end: '2016'
     ftime_max: 6
diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index 316b26f7..8f945b6c 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -139,16 +139,21 @@ check_recipe <- function(recipe) {
       error_status <- T
     }
     ## TODO: Implement multiple regions
-    # nregions <- length(recipe$Analysis$Region$Regional)
+    # nregions <- length(recipe$Analysis$Region)
     # for (i in 1:length(recipe$Analysis$Region)) {
     #   if (!all(limits %in% names(recipe$Analysis$Region[[i]]))) {
     #     limits <- paste(limits, collapse = " ")
     #     error(recipe$Run$logger,
-    #        paste("Each region defined in element 'Regional'",
-    #              "should have 4 elements:",
-    #              limits))
-    #     stop("EXECUTION FAILED")
+    #           paste0("Each region defined in element 'Region' ",
+    #                  "should have 4 elements: ",
+    #                  paste(limits, collapse = ", "), "."))
+    #     error_status <- T
     #   }
+    #   if (length(recipe$Analysis$Region) > 1) {
+    #     if (!("name" %in% names(recipe$Analysis$Region[[i]]))) {
+    #       error(recipe$Run$logger,
+    #             paste("If multiple regions are requested, each region must",
+    #                   "have a 'name'".)
     #   # are numeric? class list mode list 
     # }
   } else {
@@ -182,7 +187,7 @@ check_recipe <- function(recipe) {
   if (!is.character(recipe$Run$code_dir)) {
     error(recipe$Run$logger,
           paste("The Run element 'code_dir' in", recipe$name, "file ",
-                "should be a character string indicating the path ",
+                "should be a character string indicating the path",
                 "where the code is."))
     error_status <- T
   }
@@ -197,8 +202,8 @@ check_recipe <- function(recipe) {
   if (!is.character(recipe$Run$Loglevel) || 
       !any(recipe$Run$Loglevel %in% LOG_LEVELS)) {
     error(logger,
-          paste("The Run element 'Loglevel' in", recipe$name, "file ",
-                "should be a character string specifying one of the levels available: ",
+          paste("The Run element 'Loglevel' in", recipe$name, "file",
+                "should be a character string specifying one of the levels available:",
                 paste0(LOG_LEVELS, collapse='/')))
     error_status <- T
   }
@@ -208,16 +213,17 @@ check_recipe <- function(recipe) {
   # ---------------------------------------------------------------------
   # Check workflow: need to define restrictions? 
   # e.g. only one calibration method
-  ## TODO: Implement numbr of dependent verifications
+  ## TODO: Implement number of dependent verifications
   #nverifications <- check_number_of_dependent_verifications(recipe)
   info(recipe$Run$logger, paste("Start Dates:",
 				paste(fcst.sdate, collapse = " ")))
 
   # Return error if any check has failed
   if (error_status) {
-    stop("Recipe check failed. Please check the logs for a list of errors.")
+    error(recipe$Run$logger, "RECIPE CHECK FAILED.")
+    stop("The recipe contains some errors. The full list is in the logs.")
   } else {
-    info(recipe$Run$logger, "##### RECIPE CHECKS WERE SUCCESSFULL #####.")
+    info(recipe$Run$logger, "##### RECIPE CHECK SUCCESSFULL #####")
     # return(append(nverifications, fcst.sdate))
   }
 }
-- 
GitLab


From d3193f9aaf3a4dba672b277f36befd40ce6e022e Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 20 Dec 2022 16:07:51 +0100
Subject: [PATCH 07/21] update checker

---
 tools/check_recipe.R | 104 ++++++++++++++++++++++++++++++-------------
 1 file changed, 73 insertions(+), 31 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index 8f945b6c..a70a5b90 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -46,12 +46,13 @@ check_recipe <- function(recipe) {
   }
   ## TODO: Is this needed?
   if (is.null(recipe$Analysis$Time$fcst_year) || 
-      recipe$Analysis$Time$fcst_year == 'None') {
+      tolower(recipe$Analysis$Time$fcst_year) == 'none') {
     stream <- "hindcast"
     # recipe$Analysis$Time$fcst_year <- 'YYYY'
   } else {
     stream <- "fcst"
   }
+
   ## TODO: To be implemented in the future
   # if (length(recipe$Analysis$Time$sdate$fcst_day) > 1 &&
   #     tolower(recipe$Analysis$Horizon) != "subseasonal") {
@@ -65,6 +66,7 @@ check_recipe <- function(recipe) {
   #   error(recipe$Run$logger,
   #         paste("The element 'fcst_sday' in the recipe should be defined."))
   # }
+
   if (is.null(recipe$Analysis$Time$fcst_year)) {
     warn(recipe$Run$logger,
          paste("The element 'fcst_year' is not defined in the recipe.",
@@ -88,12 +90,13 @@ check_recipe <- function(recipe) {
   }
   # TODO: Add Workflow checks?
   # ...
-  # calculate number of workflows to create for each variable and 
+  # calculate number of workflows to create for each variable and
   if (length(recipe$Analysis$Horizon) > 1) {
     error(recipe$Run$logger,
 	  "Only one single Horizon can be specified in the recipe")
     error_status <- T
   }
+
   ## TODO: Refine this
   # nvar <- length(recipe$Analysis$Variables)
   # if (nvar > 2) {
@@ -101,6 +104,7 @@ check_recipe <- function(recipe) {
   #         "Only two type of Variables can be listed: ECVs and Indicators.")
   #   stop("EXECUTION FAILED")
   # }
+
   # remove NULL or None Indicators or ECVs from the recipe:
   if (!is.null(recipe$Analysis$Variables$Indicators) &&
       !is.list(recipe$Analysis$Variables$Indicators)) {
@@ -112,32 +116,15 @@ check_recipe <- function(recipe) {
     recipe$Analysis$Variables <- recipe$Analysis$Variables[
            -which(names(recipe$Analysis$Variables) == 'ECVs')]
   }
-  # Only one Calibration method allowed:
-  ## TODO: Make sure this condition is correctly written
-  if ((is.logical(recipe$Analysis$Workflow$Calibration$method) && 
-                  recipe$Analysis$Workflow$Calibration$method == FALSE) ||
-                  recipe$Analysis$Workflow$Calibration$method == 'None' ||
-                  is.null(recipe$Analysis$Workflow$Calibration$method)) {
-    warn(recipe$Run$logger,
-         "No Calibration method was specified, raw data verification.")
-    recipe$Analysis$Workflow$Calibration$method <- 'raw'
-  } else {
-    # remove multiple calibration methods
-    if (is.null(names(recipe$Analysis$Workflow$Calibration))) {
-      error(recipe$Run$logger,
-            "The 'Calibration' element 'method' must be specified.")
-      error_status <- T
-    }
-  }
 
-  if ("Region" %in% names(recipe$Analysis)) {
-    limits <- c('latmin', 'latmax', 'lonmin', 'lonmax')
-    if (!all(limits %in% names(recipe$Analysis$Region))) {
-      error(recipe$Run$logger,
-	    paste0("There must be 4 elements in 'Region': ",
-		   paste(limits, collapse = ", "), "."))
-      error_status <- T
-    }
+  # Region checks:
+  LIMITS <- c('latmin', 'latmax', 'lonmin', 'lonmax')
+  if (!all(LIMITS %in% names(recipe$Analysis$Region))) {
+    error(recipe$Run$logger,
+          paste0("There must be 4 elements in 'Region': ",
+		 paste(LIMITS, collapse = ", "), "."))
+    error_status <- T
+  }
     ## TODO: Implement multiple regions
     # nregions <- length(recipe$Analysis$Region)
     # for (i in 1:length(recipe$Analysis$Region)) {
@@ -156,11 +143,66 @@ check_recipe <- function(recipe) {
     #                   "have a 'name'".)
     #   # are numeric? class list mode list 
     # }
+  # ---------------------------------------------------------------------
+  # WORKFLOW CHECKS
+  # ---------------------------------------------------------------------
+
+  # Only one Calibration method allowed:
+  if ((is.logical(recipe$Analysis$Workflow$Calibration$method) &&
+       recipe$Analysis$Workflow$Calibration$method == FALSE) ||
+      tolower(recipe$Analysis$Workflow$Calibration$method) == 'none' ||
+      is.null(recipe$Analysis$Workflow$Calibration$method)) {
+    warn(recipe$Run$logger,
+         "No Calibration method was specified, raw data verification.")
+    recipe$Analysis$Workflow$Calibration$method <- 'raw'
   } else {
-    error(recipe$Run$logger, "'Region' must be defined.")
+    if (is.null(recipe$Analysis$Workflow$Calibration$method)) {
+      error(recipe$Run$logger,
+            "The 'Calibration' element 'method' must be specified.")
+      error_status <- T
+    }
+  }
+  # Anomalies
+  if ("Anomalies" %in% names(recipe$Analysis$Workflow)) {
+    if (!is.null(recipe$Analysis$Workflow$Anomalies$compute)) {
+      error(recipe$Run$logger,
+	    "Parameter 'compute' must be defined under 'Anomalies'.")
+      error_status <- T
+    } else if (!(is.logical(recipe$Analysis$Workflows$Anomalies$compute))) {
+      error(recipe$Run$logger,
+	    paste("Parameter 'Anomalies:compute' must be a logical value",
+		  "(True/False or yes/no)."))
+      error_status <- T
+    } else if ((recipe$Analysis$Workflows$Anomalies$compute) &&
+	       (!is.logical(recipe$Analysis$Workflow$Anomalies))) {
+      error(recipe$Run$logger,
+	    paste("If anomaly computation is requested, parameter",
+		  "'cross_validation' must be defined under 'Anomalies',
+		  and it must be a logical value (True/False or yes/no)."))
+      error_status <- T
+    }
+  }
+  # Skill
+  if (("Skill" %in% names(recipe$Analysis$Workflow)) &&
+      (is.null(recipe$Analysis$Workflow$Skill$metric))) {
+    error(recipe$Run$logger,
+	  "Parameter 'metric' must be defined under 'Skill'.")
     error_status <- T
   }
-  
+  # Probabilities
+  if ("Probabilities" %in% names(recipe$Analysis$Workflow)) {
+    if (is.null(recipe$Analysis$Workflow$Probabilities$percentiles)) {
+      error(recipe$Run$logger,
+	    "Parameter 'percentiles' must be defined under 'Skill'.")
+      error_status <- T
+    } else if (!is.list(recipe$Analysis$Workflow$Probabilities$percentiles)) {
+      error(recipe$Run$logger,
+	    paste("Parameter 'Probabilities:percentiles' expects a list.",
+		  "See documentation in the wiki for examples."))
+      error_status <- T
+    }
+  }
+
   # ---------------------------------------------------------------------
   # RUN CHECKS
   # ---------------------------------------------------------------------
@@ -215,8 +257,8 @@ check_recipe <- function(recipe) {
   # e.g. only one calibration method
   ## TODO: Implement number of dependent verifications
   #nverifications <- check_number_of_dependent_verifications(recipe)
-  info(recipe$Run$logger, paste("Start Dates:",
-				paste(fcst.sdate, collapse = " ")))
+  # info(recipe$Run$logger, paste("Start Dates:",
+  #  				  paste(fcst.sdate, collapse = " ")))
 
   # Return error if any check has failed
   if (error_status) {
-- 
GitLab


From a20f175b79fe2910ce25851d4aa7276b537fc087 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 20 Dec 2022 16:08:15 +0100
Subject: [PATCH 08/21] Add recipe checker to prepare_outputs

---
 modules/test_seasonal.R |  2 +-
 tools/prepare_outputs.R | 22 ++++++++++++----------
 2 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/modules/test_seasonal.R b/modules/test_seasonal.R
index 854e293f..c74fd1b1 100644
--- a/modules/test_seasonal.R
+++ b/modules/test_seasonal.R
@@ -6,7 +6,7 @@ source("modules/Saving/Saving.R")
 source("modules/Visualization/Visualization.R")
 
 recipe_file <- "modules/Loading/testing_recipes/recipe_seasonal-tests.yml"
-recipe <- prepare_outputs(recipe_file)
+recipe <- prepare_outputs(recipe_file, enable_checks = F)
 ## archive <- read_yaml(paste0(recipe$Run$code_dir, "conf/archive.yml"))$archive
 
 # Load datasets
diff --git a/tools/prepare_outputs.R b/tools/prepare_outputs.R
index da5d2562..d6eb970e 100644
--- a/tools/prepare_outputs.R
+++ b/tools/prepare_outputs.R
@@ -20,10 +20,11 @@
 #'
 #'@export
 
-prepare_outputs <- function(recipe_file) {
+prepare_outputs <- function(recipe_file,
+			    enable_checks = TRUE) {
 
-# recipe: the content of the readed recipe
-# file: the recipe file name
+# recipe_file: path to recipe YAML file
+# enable_checker: Whether or not to run the recipe checker
 
   recipe <- read_yaml(recipe_file)
   recipe$recipe_path <- recipe_file
@@ -33,32 +34,30 @@ prepare_outputs <- function(recipe_file) {
   # Create output folders:
   folder_name <- paste0(gsub(".yml", "", gsub("/", "_", recipe$name)), "_",
                         gsub(" ", "", gsub(":", "", gsub("-", "", Sys.time()))))
-
   print("Saving all outputs to:")
   print(output_dir)
   print(folder_name)
-
   dir.create(file.path(output_dir, folder_name, 'outputs'), recursive = TRUE)
   dir.create(file.path(output_dir, folder_name, 'logs'))
   dir.create(file.path(output_dir, folder_name, 'logs', 'recipes'))
-
+  # Copy recipe to output folder
   file.copy(recipe$recipe_path, file.path(output_dir, folder_name, 'logs',
 				          'recipes'))
-
+  # Create log output file
   logfile <- file.path(output_dir, folder_name, 'logs', 'log.txt')
   file.create(logfile)
 
-  # Set default behaviour of log output file:
+  # Set default behaviour of logger
   if (is.null(recipe$Run)) {
     recipe$Run <- list(Loglevel = 'INFO', Terminal = TRUE)
   }
   if (is.null(recipe$Run$Loglevel)) {
     recipe$Run$Loglevel <- 'INFO'
   }
-
   if (!is.logical(recipe$Run$Terminal)) {
     recipe$Run$Terminal <- TRUE
   }
+  # logger set-up
   if (recipe$Run$Terminal) {
     logger <- log4r::logger(threshold = recipe$Run$Loglevel,  
                             appenders = list(console_appender(layout = default_log_layout()),
@@ -69,10 +68,13 @@ prepare_outputs <- function(recipe_file) {
                             appenders = list(file_appende(logfile, append = TRUE,             
                                              layout = default_log_layout())))
   }
-
   recipe$Run$output_dir <- file.path(output_dir, folder_name)
   recipe$Run$logger <- logger
   recipe$Run$logfile <- logfile
+  # Run recipe checker
+  if (enable_checks) {
+    check_recipe(recipe)
+  }
 
   return(recipe)
 }
-- 
GitLab


From ba419683e13cc1d14217b0116edabbb22bf317e2 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 20 Dec 2022 16:18:58 +0100
Subject: [PATCH 09/21] Print full path to recipe; fix bug in Anomalies check

---
 tools/check_recipe.R | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index a70a5b90..c17a7cd9 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -3,7 +3,7 @@ check_recipe <- function(recipe) {
   # recipe: yaml recipe already read it
   # output: errors or the total number of workflow (vars x regions) to compute
 
-  info(recipe$Run$logger, paste("Checking recipe:", recipe$name))
+  info(recipe$Run$logger, paste("Checking recipe:", recipe$path))
 
   # ---------------------------------------------------------------------
   # ANALYSIS CHECKS
@@ -164,7 +164,7 @@ check_recipe <- function(recipe) {
   }
   # Anomalies
   if ("Anomalies" %in% names(recipe$Analysis$Workflow)) {
-    if (!is.null(recipe$Analysis$Workflow$Anomalies$compute)) {
+    if (is.null(recipe$Analysis$Workflow$Anomalies$compute)) {
       error(recipe$Run$logger,
 	    "Parameter 'compute' must be defined under 'Anomalies'.")
       error_status <- T
-- 
GitLab


From 89ed146257cbded2fb8e7cc52cad5c54a0bc07d5 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 20 Dec 2022 16:19:48 +0100
Subject: [PATCH 10/21] Add TODO

---
 tools/check_recipe.R | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index c17a7cd9..2381309b 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -1,7 +1,7 @@
 check_recipe <- function(recipe) {
 
   # recipe: yaml recipe already read it
-  # output: errors or the total number of workflow (vars x regions) to compute
+  ## TODO: Adapt to decadal case
 
   info(recipe$Run$logger, paste("Checking recipe:", recipe$path))
 
-- 
GitLab


From 508683900950925f1bb894a447f49d2cdc34e841 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 20 Dec 2022 16:22:02 +0100
Subject: [PATCH 11/21] Fix bug caused by typo

---
 tools/check_recipe.R | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index 2381309b..153c4a18 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -168,12 +168,12 @@ check_recipe <- function(recipe) {
       error(recipe$Run$logger,
 	    "Parameter 'compute' must be defined under 'Anomalies'.")
       error_status <- T
-    } else if (!(is.logical(recipe$Analysis$Workflows$Anomalies$compute))) {
+    } else if (!(is.logical(recipe$Analysis$Workflow$Anomalies$compute))) {
       error(recipe$Run$logger,
 	    paste("Parameter 'Anomalies:compute' must be a logical value",
 		  "(True/False or yes/no)."))
       error_status <- T
-    } else if ((recipe$Analysis$Workflows$Anomalies$compute) &&
+    } else if ((recipe$Analysis$Workflow$Anomalies$compute) &&
 	       (!is.logical(recipe$Analysis$Workflow$Anomalies))) {
       error(recipe$Run$logger,
 	    paste("If anomaly computation is requested, parameter",
-- 
GitLab


From eb36caeeb545ce7811d2f8d2fb239f59b6f71a0a Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 20 Dec 2022 16:35:37 +0100
Subject: [PATCH 12/21] Fix pipeline

---
 tools/check_recipe.R | 33 +++++++++++++++++++++++----------
 1 file changed, 23 insertions(+), 10 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index 153c4a18..fcd30ef3 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -3,16 +3,18 @@ check_recipe <- function(recipe) {
   # recipe: yaml recipe already read it
   ## TODO: Adapt to decadal case
 
-  info(recipe$Run$logger, paste("Checking recipe:", recipe$path))
+  info(recipe$Run$logger, paste("Checking recipe:", recipe$recipe_path))
 
   # ---------------------------------------------------------------------
   # ANALYSIS CHECKS
   # ---------------------------------------------------------------------
 
-  TIME_SETTINGS = c('sdate', 'ftime_min', 'ftime_max', 'hcst_start', 'hcst_end')
-  PARAMS = c('Horizon', 'Time', 'Variables', 'Region', 'Regrid', 'Workflow',
-	     'Datasets')
-  HORIZONS <- c('subseasonal', 'seasonal', 'decadal')
+  TIME_SETTINGS_SEASONAL <- c("sdate", "ftime_min", "ftime_max", "hcst_start",
+			      "hcst_end")
+  TIME_SETTINGS_DECADAL <- c("ftime_min", "ftime_max", "hcst_start", "hcst_end") 
+  PARAMS <- c("Horizon", "Time", "Variables", "Region", "Regrid", "Workflow",
+	      "Datasets")
+  HORIZONS <- c("subseasonal", "seasonal", "decadal")
   # Define error status variable
   error_status <- F
 
@@ -38,11 +40,22 @@ check_recipe <- function(recipe) {
   }
   # Check temporal settings and
   # count the number of verifications
-  if (!all(TIME_SETTINGS %in% names(recipe$Analysis$Time))) {
-    error(recipe$Run$logger,
-	  paste0("The element 'Time' in the recipe must contain all of the ",
-		 "following: ", paste(TIME_SETTINGS, collapse = ", "), "."))
-    error_status <- T
+  if (tolower(recipe$Analysis$Horizon) == "seasonal") {
+    if (!all(TIME_SETTINGS_SEASONAL %in% names(recipe$Analysis$Time))) {
+      error(recipe$Run$logger,
+	    paste0("The element 'Time' in the recipe must contain all of the ",
+		   "following: ", paste(TIME_SETTINGS_SEASONAL,
+					collapse = ", "), "."))
+      error_status <- T
+    }
+  } else if (tolower(recipe$Analysis$Horizon) == "decadal") {
+    if (!all(TIME_SETTINGS_DECADAL %in% names(recipe$Analysis$Time))) {
+      error(recipe$Run$logger,
+	    paste0("The element 'Time' in the recipe must contain all of the ",
+		   "following: ", paste(TIME_SETTINGS_DECADAL,
+					collapse = ", "), "."))
+      error_status <- T
+    }
   }
   ## TODO: Is this needed?
   if (is.null(recipe$Analysis$Time$fcst_year) || 
-- 
GitLab


From 7a985badf5d0cb06eec6504d56f20c1100ca1d57 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Fri, 23 Dec 2022 15:01:01 +0100
Subject: [PATCH 13/21] Add more detailed time section checks

---
 tools/check_recipe.R | 39 +++++++++++++++++++++++++++++++++++++--
 1 file changed, 37 insertions(+), 2 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index fcd30ef3..5fca3597 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -38,8 +38,7 @@ check_recipe <- function(recipe) {
 		"following: ", paste(HORIZONS, collapse = ", "), "."))
     error_status <- T
   }
-  # Check temporal settings and
-  # count the number of verifications
+  # Check time settings
   if (tolower(recipe$Analysis$Horizon) == "seasonal") {
     if (!all(TIME_SETTINGS_SEASONAL %in% names(recipe$Analysis$Time))) {
       error(recipe$Run$logger,
@@ -57,6 +56,42 @@ check_recipe <- function(recipe) {
       error_status <- T
     }
   }
+  # Check ftime_min and ftime_max
+  if ((!(recipe$Analysis$Time$ftime_min > 0)) ||
+      (!is.integer(recipe$Analysis$Time$ftime_min))) {
+    error(recipe$Run$logger,
+	  "The element 'ftime_min' must be an integer larger than 0.")
+    error_status <- T
+  }
+  if ((!(recipe$Analysis$Time$ftime_max > 0)) ||
+      (!is.integer(recipe$Analysis$Time$ftime_max))) {
+    error(recipe$Run$logger,
+	  "The element 'ftime_max' must be an integer larger than 0.")
+    error_status <- T
+  }
+  if (recipe$Analysis$Time$ftime_max < recipe$Analysis$Time$ftime_min) {
+    error(recipe$Run$logger,
+	  "'ftime_max' cannot be smaller than 'ftime_min'.")
+    error_status <- T
+  }
+  # Check consistency of hindcast years
+  if (!(as.numeric(recipe$Analysis$Time$hcst_start) %% 1 == 0) ||
+      (!(recipe$Analysis$Time$hcst_start > 0))) {
+    error(recipe$Run$logger,
+	  "The element 'hcst_start' must be a valid year.")
+    error_status <- T
+  }
+  if (!(as.numeric(recipe$Analysis$Time$hcst_end) %% 1 == 0) ||
+      (!(recipe$Analysis$Time$hcst_end > 0))) {
+    error(recipe$Run$logger,
+	  "The element 'hcst_end' must be a valid year.")
+    error_status <- T
+  }
+  if (recipe$Analysis$Time$hcst_end < recipe$Analysis$Time$hcst_start) {
+    error(recipe$Run$logger,
+	  "'hcst_end' cannot be smaller than 'hcst_start'.")
+    error_status <- T
+  }
   ## TODO: Is this needed?
   if (is.null(recipe$Analysis$Time$fcst_year) || 
       tolower(recipe$Analysis$Time$fcst_year) == 'none') {
-- 
GitLab


From d7ea2794b6df6e1492888546679aeba8758f63e3 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Thu, 29 Dec 2022 15:11:47 +0100
Subject: [PATCH 14/21] Update script to divide recipe

---
 ...ample.yml => recipe_splitting_example.yml} | 16 +--
 tools/divide_recipe.R                         | 98 ++++++++++---------
 2 files changed, 61 insertions(+), 53 deletions(-)
 rename recipes/{recipe_example.yml => recipe_splitting_example.yml} (86%)

diff --git a/recipes/recipe_example.yml b/recipes/recipe_splitting_example.yml
similarity index 86%
rename from recipes/recipe_example.yml
rename to recipes/recipe_splitting_example.yml
index 27f74df9..e62611ab 100644
--- a/recipes/recipe_example.yml
+++ b/recipes/recipe_splitting_example.yml
@@ -4,7 +4,7 @@
 
 Description:
   Author: V. Agudetse
-  Info: Test for main recipe
+  Info: Test for recipe splitting
 
 ################################################################################
 ## ANALYSIS CONFIGURATION
@@ -21,23 +21,25 @@ Analysis:
       - {name: system5c3s}
     Multimodel: False # single option
     Reference:
-      name: era5 # multiple references for single model?
+      - {name: era5} # multiple references for single model?
   Time:
-    sdate: '1101' # list, split
+    sdate: # list, split
+      - '1101'
+      - '1201'
     fcst_year: '2020' # list, don't split, handled internally
     hcst_start: '1993' # single option
     hcst_end: '2016' # single option
     ftime_min: 1 # single option
-    ftime_max: 6 # singl eoption
+    ftime_max: 6 # single option
   Region: # multiple lists, split? Add region name if length(Region) > 1
-    - {name: "global", latmin: -90, latmax: 90, lonmin: 0, lonmax: 360}
+    - {name: "global", latmin: -90, latmax: 90, lonmin: 0, lonmax: 359.9}
     - {name: "nino34", latmin: -5, latmax: 5, lonmin: -10, lonmax: 60}
   Regrid:
-    method: bilinear # allow multiple methods?
+    method: bilinear ## TODO: allow multiple methods?
     type: to_system
   Workflow:
     Calibration:
-      method: mse_min # list, split?
+      method: mse_min ## TODO: list, split?
     Skill:
       metric: RPS, RPSS, CRPS, CRPSS, FRPSS, BSS10, BSS90, mean_bias, mean_bias_SS # list, don't split
     Probabilities:
diff --git a/tools/divide_recipe.R b/tools/divide_recipe.R
index 8f4ef35a..fd4efc6e 100644
--- a/tools/divide_recipe.R
+++ b/tools/divide_recipe.R
@@ -1,11 +1,8 @@
-# recipe: the content of the recipe
-# verifications: the output from check_recipe
-# folder: the name of the output folder for this run
-# logger: the log file obtain from prepare_outputs
+# recipe: the recipe as returned by prepare_outputs()
+divide_recipe <- function(recipe) {
 
-## TODO: Rethink strategy and adapt function accordingly
-divide_recipe <- function(recipe, verifications, folder, logger) {
-  info(logger, "Spliting recipe in single verifications.")
+  ## TODO: Implement dependent vs independent verifications?
+  info(recipe$Run$logger, "Spliting recipe in single verifications.")
   beta_recipe <- list(Description = append(recipe$Description, 
                                            "split version"),
                       Analysis = list(Horizon = recipe$Analysis$Horizon,
@@ -17,37 +14,37 @@ divide_recipe <- function(recipe, verifications, folder, logger) {
                                       Workflow = recipe$Analysis$Workflow,
                                       Output_format = 
 				        recipe$Analysis$Output_format),
-		      Run = recipe$Run)
-  # duplicate recipe by Variables considering dep and indep:
-  all_recipes <- list(beta_recipe)
-  i <- 1 # to get track of the recipe number
-  for (indep in verifications$independent) {
-    all_recipes[[i]]$Analysis$Variables <- indep
-    i = i + 1
-    all_recipes <- append(all_recipes, list(beta_recipe))
-  }
-  for (dep in verifications$dependent) {
-    all_recipes[[i]]$Analysis$Variables <- dep
-    i = i + 1
-    all_recipes <- append(all_recipes, list(beta_recipe))
+		      Run = recipe$Run[c("Loglevel", "output_dir", "Terminal",
+					 "code_dir", "logfile")])
+
+  # duplicate recipe by independent variables:
+  all_recipes <- rep(list(beta_recipe), length(recipe$Analysis$Variables))
+  for (var in 1:length(recipe$Analysis$Variables)) {
+    all_recipes[[var]]$Analysis$Variables <- recipe$Analysis$Variables[[var]]
   }
-  all_recipes <- all_recipes[-length(all_recipes)]
+  # for (dep in verifications$dependent) {
+  #   all_recipes[[i]]$Analysis$Variables <- dep
+  #   i = i + 1
+  #   all_recipes <- append(all_recipes, list(beta_recipe))
+  # }
+  # all_recipes <- all_recipes[-length(all_recipes)] # wth does this do
+  
   # duplicate recipe by Datasets:
   # check Systems
   if (recipe$Analysis$Datasets$Multimodel) {
     for (reci in 1:length(all_recipes)) {
-      all_recipes[[reci]]$Analysis$Datasets <- list(
-                            System = recipe$Analysis$Datasets$System,
-                            Multimodel = recipe$Analysis$Datasets$Multimodel,
-                            Reference = NULL)
+      all_recipes[[reci]]$Analysis$Datasets <- 
+	list(System = recipe$Analysis$Datasets$System,
+	     Multimodel = recipe$Analysis$Datasets$Multimodel,
+	     Reference = NULL)
     }
   } else {
     for (sys in 1:length(recipe$Analysis$Datasets$System)) {
       for (reci in 1:length(all_recipes)) {
-        all_recipes[[reci]]$Analysis$Datasets <- list(
-                            System = recipe$Analysis$Datasets$System[[sys]],
-                            Multimodel = recipe$Analysis$Datasets$Multimodel,
-                            Reference = NULL)
+        all_recipes[[reci]]$Analysis$Datasets <- 
+	  list(System = recipe$Analysis$Datasets$System[[sys]],
+	       Multimodel = recipe$Analysis$Datasets$Multimodel,
+	       Reference = NULL)
       }
       if (sys == 1) {
         recipes <- all_recipes
@@ -74,28 +71,28 @@ divide_recipe <- function(recipe, verifications, folder, logger) {
   # Duplicate recipe by Region
   recipes <- list()
   for (reg in 1:length(recipe$Analysis$Region)) {
-    if (length(recipe$Analysis$Region[[reg]]) == 4) { ##TODO: THIS SHOULD BE ONLY CHECK IN THE RECIPE CHECKER?
+    # if (length(recipe$Analysis$Region[[reg]]) == 4) { ##TODO: THIS SHOULD BE ONLY CHECK IN THE RECIPE CHECKER?
       for (reci in 1:length(all_recipes)) {
-        all_recipes[[reci]]$Analysis$Region <-
-                              recipe$Analysis$Region[[reg]]
+        all_recipes[[reci]]$Analysis$Region <- recipe$Analysis$Region[[reg]]
       }
       recipes <- append(recipes, all_recipes)
-    }
+    # }
   }
   all_recipes <- recipes
   rm(list = 'recipes')
+  # Duplicate recipe by start date
   if (tolower(recipe$Analysis$Horizon) == 'seasonal') {
-    for (sday in 1:length(recipe$Analysis$Time$sdate$fcst_sday)) {
+    for (sdate in 1:length(recipe$Analysis$Time$sdate)) {
       for (reci in 1:length(all_recipes)) {
-      all_recipes[[reci]]$Analysis$Time <- list(sdate = list(
-                      fcst_syear = recipe$Analysis$Time$sdate$fcst_syear,
-                      fcst_sday = recipe$Analysis$Time$sdate$fcst_sday[[sday]]),
-                      hcst_start = recipe$Analysis$Time$hcst_start,
-                      hcst_end = recipe$Analysis$Time$hcst_end,
-                      leadtimemin = recipe$Analysis$Time$leadtimemin,
-                      leadtimemax = recipe$Analysis$Time$leadtimemax)
+      all_recipes[[reci]]$Analysis$Time <-
+	list(sdate = recipe$Analysis$Time$sdate[[sdate]],
+	     fcst_year = recipe$Analysis$Time$fcst_year,
+	     hcst_start = recipe$Analysis$Time$hcst_start,
+	     hcst_end = recipe$Analysis$Time$hcst_end,
+	     ftime_min = recipe$Analysis$Time$ftime_min,
+	     ftime_max = recipe$Analysis$Time$ftime_max)
       }
-      if (sday == 1) {
+      if (sdate == 1) {
         recipes <- all_recipes
       } else {
         recipes <- append(recipes, all_recipes)
@@ -104,12 +101,21 @@ divide_recipe <- function(recipe, verifications, folder, logger) {
     all_recipes <- recipes
     rm(list = 'recipes')
   } # Rest of horizons
-  # Finally, save all recipes in saparated yaml files
+  # Save all recipes in separate YAML files
+  ## TODO: Re-add recipe$Run$logger
   for (reci in 1:length(all_recipes)) {
+    if (reci < 10) {
+      recipe_number <- paste0("0", reci)
+    } else {
+      recipe_number <- reci
+    }
     write_yaml(all_recipes[[reci]], 
-               paste0(folder, "/logs/recipes/recipe_", reci, ".yml"))
+               paste0(recipe$Run$output_dir, "/logs/recipes/recipe_",
+		      recipe_number, ".yml"))
+    all_recipes[[reci]]$Run$logger <- recipe$Run$logger
   }
-  text <- paste0("See folder ",folder,"/logs/recipes/ to see the individual recipes.")
-  info(logger, text)
+  text <- paste0("See recipe$Run$output_dir ", recipe$Run$output_dir,
+		 "/logs/recipes/ to see the individual recipes.")
+  info(recipe$Run$logger, text)
   return(all_recipes)
 }
-- 
GitLab


From 449258ad087a30faa1ae236eb381bd4e711dd481 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Thu, 29 Dec 2022 15:14:25 +0100
Subject: [PATCH 15/21] Move old test recipes to new directory

---
 recipes/tests/{ => old_tests}/execute_tests.R            | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow1.yml | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow2.yml | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow3.yml | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow4.yml | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow5.yml | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow6.yml | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow7.yml | 0
 recipes/tests/{ => old_tests}/seasonal_testWorkflow8.yml | 0
 9 files changed, 0 insertions(+), 0 deletions(-)
 rename recipes/tests/{ => old_tests}/execute_tests.R (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow1.yml (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow2.yml (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow3.yml (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow4.yml (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow5.yml (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow6.yml (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow7.yml (100%)
 rename recipes/tests/{ => old_tests}/seasonal_testWorkflow8.yml (100%)

diff --git a/recipes/tests/execute_tests.R b/recipes/tests/old_tests/execute_tests.R
similarity index 100%
rename from recipes/tests/execute_tests.R
rename to recipes/tests/old_tests/execute_tests.R
diff --git a/recipes/tests/seasonal_testWorkflow1.yml b/recipes/tests/old_tests/seasonal_testWorkflow1.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow1.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow1.yml
diff --git a/recipes/tests/seasonal_testWorkflow2.yml b/recipes/tests/old_tests/seasonal_testWorkflow2.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow2.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow2.yml
diff --git a/recipes/tests/seasonal_testWorkflow3.yml b/recipes/tests/old_tests/seasonal_testWorkflow3.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow3.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow3.yml
diff --git a/recipes/tests/seasonal_testWorkflow4.yml b/recipes/tests/old_tests/seasonal_testWorkflow4.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow4.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow4.yml
diff --git a/recipes/tests/seasonal_testWorkflow5.yml b/recipes/tests/old_tests/seasonal_testWorkflow5.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow5.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow5.yml
diff --git a/recipes/tests/seasonal_testWorkflow6.yml b/recipes/tests/old_tests/seasonal_testWorkflow6.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow6.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow6.yml
diff --git a/recipes/tests/seasonal_testWorkflow7.yml b/recipes/tests/old_tests/seasonal_testWorkflow7.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow7.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow7.yml
diff --git a/recipes/tests/seasonal_testWorkflow8.yml b/recipes/tests/old_tests/seasonal_testWorkflow8.yml
similarity index 100%
rename from recipes/tests/seasonal_testWorkflow8.yml
rename to recipes/tests/old_tests/seasonal_testWorkflow8.yml
-- 
GitLab


From db4d66fb1edb19085f882ed545df2b66740de349 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Thu, 29 Dec 2022 15:18:56 +0100
Subject: [PATCH 16/21] Change enable_checks param to disable_checks, default
 FALSE

---
 tools/prepare_outputs.R | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/tools/prepare_outputs.R b/tools/prepare_outputs.R
index d6eb970e..56ab56c6 100644
--- a/tools/prepare_outputs.R
+++ b/tools/prepare_outputs.R
@@ -1,7 +1,7 @@
 #'Read recipe YAML file and create and store logfile info
 #'
 #'The purpose of this function is to read the recipe configuration for Auto-S2S
-#'workflows and create logfiles stores in an the output directory specified in 
+#'workflows and create logfiles stored in an the output directory specified in 
 #'the recipe. It returns an object of class logger that stores information on 
 #'the recipe configuration and errors.
 #'
@@ -21,10 +21,10 @@
 #'@export
 
 prepare_outputs <- function(recipe_file,
-			    enable_checks = TRUE) {
+			    disable_checks = FALSE) {
 
 # recipe_file: path to recipe YAML file
-# enable_checker: Whether or not to run the recipe checker
+# disable_checks: If TRUE, does not perform checks on recipe
 
   recipe <- read_yaml(recipe_file)
   recipe$recipe_path <- recipe_file
@@ -72,9 +72,8 @@ prepare_outputs <- function(recipe_file,
   recipe$Run$logger <- logger
   recipe$Run$logfile <- logfile
   # Run recipe checker
-  if (enable_checks) {
+  if !(disable_checks) {
     check_recipe(recipe)
   }
-
   return(recipe)
 }
-- 
GitLab


From 9187ec11aa61e6ad87d0f257ffb3e593328e832f Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Thu, 29 Dec 2022 16:36:28 +0100
Subject: [PATCH 17/21] New recipe; test multiple verifications run serially;
 fix bug in prepare_outputs()

---
 modules/test_seasonal.R                       |  3 +-
 .../tests/recipe_seasonal_two-variables.yml   | 60 +++++++++++++++++++
 tools/divide_recipe.R                         |  7 ++-
 tools/prepare_outputs.R                       |  2 +-
 4 files changed, 68 insertions(+), 4 deletions(-)
 create mode 100644 recipes/tests/recipe_seasonal_two-variables.yml

diff --git a/modules/test_seasonal.R b/modules/test_seasonal.R
index c74fd1b1..d776609d 100644
--- a/modules/test_seasonal.R
+++ b/modules/test_seasonal.R
@@ -5,8 +5,9 @@ source("modules/Skill/Skill.R")
 source("modules/Saving/Saving.R")
 source("modules/Visualization/Visualization.R")
 
-recipe_file <- "modules/Loading/testing_recipes/recipe_seasonal-tests.yml"
+recipe_file <- "recipes/recipe_splitting_example.yml"
 recipe <- prepare_outputs(recipe_file, enable_checks = F)
+atomic_recipes <- divide_recipe(recipe)
 ## archive <- read_yaml(paste0(recipe$Run$code_dir, "conf/archive.yml"))$archive
 
 # Load datasets
diff --git a/recipes/tests/recipe_seasonal_two-variables.yml b/recipes/tests/recipe_seasonal_two-variables.yml
new file mode 100644
index 00000000..89406ece
--- /dev/null
+++ b/recipes/tests/recipe_seasonal_two-variables.yml
@@ -0,0 +1,60 @@
+################################################################################ 
+## RECIPE DESCRIPTION  
+################################################################################ 
+
+Description:
+  Author: V. Agudetse
+  Info: Test Independent verification of two variables
+
+################################################################################
+## ANALYSIS CONFIGURATION
+################################################################################
+
+Analysis:
+  Horizon: Seasonal
+  Variables: # ECVs and Indicators?
+    - {name: tas, freq: monthly_mean}
+    - {name: prlr, freq: monthly_mean}
+  Datasets:
+    System: # multiple systems for single model, split if Multimodel = F
+      - {name: system5c3s}
+    Multimodel: False # single option
+    Reference:
+      - {name: era5} # multiple references for single model?
+  Time:
+    sdate: # list, split
+      - '0101'
+    fcst_year: '2020' # list, don't split, handled internally
+    hcst_start: '2000' # single option
+    hcst_end: '2016' # single option
+    ftime_min: 1 # single option
+    ftime_max: 3 # single option
+  Region: # multiple lists, split? Add region name if length(Region) > 1
+    - {name: "nino34", latmin: -5, latmax: 5, lonmin: -10, lonmax: 60}
+  Regrid:
+    method: bilinear ## TODO: allow multiple methods?
+    type: to_system
+  Workflow:
+    Anomalies:
+      compute: yes
+      cross_validation: yes
+    Calibration:
+      method: mse_min ## TODO: list, split?
+    Skill:
+      metric: RPS, RPSS, CRPS, CRPSS, FRPSS, BSS10, BSS90, mean_bias, mean_bias_SS # list, don't split
+    Probabilities:
+      percentiles: [[1/3, 2/3], [1/10, 9/10], [1/4, 2/4, 3/4]] # list, don't split
+    Indicators:
+      index: no # ? 
+  ncores: 7
+  remove_NAs: yes # bool, don't split
+  Output_format: S2S4E # string, don't split
+
+################################################################################
+## Run CONFIGURATION
+################################################################################
+Run:
+  Loglevel: INFO
+  Terminal: yes
+  output_dir: /esarchive/scratch/vagudets/repos/auto-s2s/out-logs/
+  code_dir: /esarchive/scratch/vagudets/repos/auto-s2s/
diff --git a/tools/divide_recipe.R b/tools/divide_recipe.R
index fd4efc6e..6cc8ce4a 100644
--- a/tools/divide_recipe.R
+++ b/tools/divide_recipe.R
@@ -114,8 +114,11 @@ divide_recipe <- function(recipe) {
 		      recipe_number, ".yml"))
     all_recipes[[reci]]$Run$logger <- recipe$Run$logger
   }
-  text <- paste0("See recipe$Run$output_dir ", recipe$Run$output_dir,
-		 "/logs/recipes/ to see the individual recipes.")
+  info(recipe$Run$logger,
+       paste("The main recipe has been devided into", length(all_recipes),
+	     "atomic recipes."))
+  text <- paste0("See output directory ", recipe$Run$output_dir,
+		 "/logs/recipes/ to see all the individual atomic recipes.")
   info(recipe$Run$logger, text)
   return(all_recipes)
 }
diff --git a/tools/prepare_outputs.R b/tools/prepare_outputs.R
index 56ab56c6..72fc5433 100644
--- a/tools/prepare_outputs.R
+++ b/tools/prepare_outputs.R
@@ -72,7 +72,7 @@ prepare_outputs <- function(recipe_file,
   recipe$Run$logger <- logger
   recipe$Run$logfile <- logfile
   # Run recipe checker
-  if !(disable_checks) {
+  if (!disable_checks) {
     check_recipe(recipe)
   }
   return(recipe)
-- 
GitLab


From 8b30132f06d7cce5b4202d68400ee36b30e382a5 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Thu, 29 Dec 2022 17:01:50 +0100
Subject: [PATCH 18/21] Add info to atomic recipe, fix typo

---
 tools/divide_recipe.R | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/tools/divide_recipe.R b/tools/divide_recipe.R
index 6cc8ce4a..18962f93 100644
--- a/tools/divide_recipe.R
+++ b/tools/divide_recipe.R
@@ -4,7 +4,9 @@ divide_recipe <- function(recipe) {
   ## TODO: Implement dependent vs independent verifications?
   info(recipe$Run$logger, "Spliting recipe in single verifications.")
   beta_recipe <- list(Description = append(recipe$Description, 
-                                           "split version"),
+                                           list(Origin = paste("Atomic recipe,",
+							       "split from:", 
+							       recipe$name))),
                       Analysis = list(Horizon = recipe$Analysis$Horizon,
                                       Variables = NULL,
                                       Datasets = NULL,
@@ -115,7 +117,7 @@ divide_recipe <- function(recipe) {
     all_recipes[[reci]]$Run$logger <- recipe$Run$logger
   }
   info(recipe$Run$logger,
-       paste("The main recipe has been devided into", length(all_recipes),
+       paste("The main recipe has been divided into", length(all_recipes),
 	     "atomic recipes."))
   text <- paste0("See output directory ", recipe$Run$output_dir,
 		 "/logs/recipes/ to see all the individual atomic recipes.")
-- 
GitLab


From 8a589bfe32da3bf2e5ddeb3662d96294f4988481 Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Fri, 30 Dec 2022 10:16:56 +0100
Subject: [PATCH 19/21] Simple example using a for loop

---
 modules/test_seasonal.R | 36 +++++++++++++++++++-----------------
 1 file changed, 19 insertions(+), 17 deletions(-)

diff --git a/modules/test_seasonal.R b/modules/test_seasonal.R
index d776609d..921e8f96 100644
--- a/modules/test_seasonal.R
+++ b/modules/test_seasonal.R
@@ -5,23 +5,25 @@ source("modules/Skill/Skill.R")
 source("modules/Saving/Saving.R")
 source("modules/Visualization/Visualization.R")
 
-recipe_file <- "recipes/recipe_splitting_example.yml"
-recipe <- prepare_outputs(recipe_file, enable_checks = F)
+recipe_file <- "recipes/tests/recipe_seasonal_two-variables.yml"
+recipe <- prepare_outputs(recipe_file, disable_checks = T)
 atomic_recipes <- divide_recipe(recipe)
 ## archive <- read_yaml(paste0(recipe$Run$code_dir, "conf/archive.yml"))$archive
 
-# Load datasets
-data <- load_datasets(recipe)
-# Calibrate datasets
-calibrated_data <- calibrate_datasets(recipe, data)
-# Compute anomalies
-calibrated_data <- compute_anomalies(recipe, calibrated_data)
-# Compute skill metrics 
-skill_metrics <- compute_skill_metrics(recipe, calibrated_data)
-# Compute percentiles and probability bins
-probabilities <- compute_probabilities(recipe, calibrated_data)
-# Export all data to netCDF
-save_data(recipe, calibrated_data, skill_metrics, probabilities)
-# Plot data
-plot_data(recipe, calibrated_data, skill_metrics, probabilities,
-	  significance = T)
+for (atomic_recipe in atomic_recipes) {
+  # Load datasets
+  data <- load_datasets(atomic_recipe)
+  # Calibrate datasets
+  calibrated_data <- calibrate_datasets(atomic_recipe, data)
+  # Compute anomalies
+  calibrated_data <- compute_anomalies(atomic_recipe, calibrated_data)
+  # Compute skill metrics 
+  skill_metrics <- compute_skill_metrics(atomic_recipe, calibrated_data)
+  # Compute percentiles and probability bins
+  probabilities <- compute_probabilities(atomic_recipe, calibrated_data)
+  # Export all data to netCDF
+  save_data(atomic_recipe, calibrated_data, skill_metrics, probabilities)
+  # Plot data
+  plot_data(atomic_recipe, calibrated_data, skill_metrics, probabilities,
+	    significance = T)
+}
-- 
GitLab


From 200d3eefd7b197e11701da2cc92e83da726a198d Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Fri, 30 Dec 2022 10:18:34 +0100
Subject: [PATCH 20/21] Add a TODO

---
 modules/Saving/paths2save.R | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/modules/Saving/paths2save.R b/modules/Saving/paths2save.R
index 2d6353fe..2d5a0a4e 100644
--- a/modules/Saving/paths2save.R
+++ b/modules/Saving/paths2save.R
@@ -1,4 +1,6 @@
 ## TODO: Separate by time aggregation
+## TODO: Build a default path that accounts for:
+## variable, system, reference, start date and region name
 
 get_filename <- function(dir, recipe, var, date, agg, file.type) {
   # This function builds the path of the output file based on directory,
-- 
GitLab


From a65b43cdcae8dfd49c1cf6b86873b9086fbf425d Mon Sep 17 00:00:00 2001
From: Victoria Agudetse Roures <victoria.agudetse@bsc.es>
Date: Tue, 31 Jan 2023 12:01:58 +0100
Subject: [PATCH 21/21] Comment lines

---
 tools/check_recipe.R | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

diff --git a/tools/check_recipe.R b/tools/check_recipe.R
index 307bc3eb..541319f5 100644
--- a/tools/check_recipe.R
+++ b/tools/check_recipe.R
@@ -124,15 +124,16 @@ check_recipe <- function(recipe) {
 	       "No forecast year will be used."))
   }
   ## TODO: Adapt and move this inside 'if'? 
-  fcst.sdate <- NULL
-  for (syear in recipe$Analysis$Time$fcst_year) {
-    for (sday in recipe$Analysis$Time$sdate) {
-      fcst.sdate <- c(fcst.sdate, 
-                      paste0(syear,
-                             sprintf("%04d", as.numeric(sday))))
-    }
-  }
-  fcst.sdate <- list(stream = stream, fcst.sdate = fcst.sdate)   
+  # fcst.sdate <- NULL
+  # for (syear in recipe$Analysis$Time$fcst_year) {
+  #   for (sday in recipe$Analysis$Time$sdate) {
+  #     fcst.sdate <- c(fcst.sdate, 
+  #                     paste0(syear,
+  #                            sprintf("%04d", as.numeric(sday))))
+  #   }
+  # }
+  # fcst.sdate <- list(stream = stream, fcst.sdate = fcst.sdate)
+
   # Regrid checks:
   if (length(recipe$Analysis$Regrid) != 2) {
     error(recipe$Run$logger,
-- 
GitLab