REMIND: Check for summation errors and report in already existing Mif column of getRunStatus

.buildlibrary

@@ -1,4 +1,4 @@
-ValidationKey: '3581447'
+ValidationKey: '3760100'
 AcceptedWarnings:
 - 'Warning: package ''.*'' was built under R version'
 - 'Warning: namespace ''.*'' is not available and has been replaced'

CITATION.cff

@@ -2,8 +2,8 @@ cff-version: 1.2.0
 message: If you use this software, please cite it using the metadata from this file.
 type: software
 title: 'modelstats: Run Analysis Tools'
-version: 0.18.1
-date-released: '2024-03-05'
+version: 0.19.0
+date-released: '2024-03-08'
 abstract: A collection of tools to analyze model runs.
 authors:
 - family-names: Giannousakis

DESCRIPTION

@@ -1,8 +1,8 @@
 Package: modelstats
 Type: Package
 Title: Run Analysis Tools
-Version: 0.18.1
-Date: 2024-03-05
+Version: 0.19.0
+Date: 2024-03-08
 Authors@R: c(
     person("Anastasis", "Giannousakis", email = "[email protected]", role = c("aut","cre")),
     person("Oliver", "Richters", role = "aut")

R/getRunStatus.R

@@ -126,8 +126,15 @@ getRunStatus <- function(mydir = dir(), sort = "nf", user = NULL) {
         miffile <- paste0(ii, "/validation.mif")
         out[i, "Mif"] <- if (file.exists(miffile) && file.info(miffile)[["size"]] > 99999) "yes" else "no"
       } else {
-        miffile <- paste0(ii, "/REMIND_generic_", cfg[["title"]], ".mif")
-        out[i, "Mif"] <- if (file.exists(miffile) && file.info(miffile)[["size"]] > 3899999) "yes" else "no"
+        miffile    <- paste0(ii, "/REMIND_generic_", cfg[["title"]], ".mif")
+        sumErrFile <- paste0(ii, "/REMIND_generic_", cfg[["title"]], "_summation_errors.csv")
+        if (! file.exists(miffile)){ 
+          out[i, "Mif"] <- "no"
+        } else if (file.exists(sumErrFile)){
+          out[i, "Mif"] <- "sumErr"
+        } else {
+          out[i, "Mif"] <- "yes"
+        }
       }
     }
 

R/loopRuns.R

@@ -43,11 +43,11 @@ loopRuns <- function(mydir, user = NULL, colors = TRUE, sortbytime = TRUE) {
   if (file.exists("/p")) {
     coltitles <- c(paste0("Folder", paste(rep(" ", len - 6), collapse = "")),
       "Runtime    ", "inSlurm ", "RunType    ", "RunStatus        ", "Iter            ",
-      "Conv                 ", "modelstat            ", "Mif", "AppResults")
+      "Conv                 ", "modelstat            ", "Mif   ", "AppResults")
     lenCols <- c(nchar(coltitles)[-length(coltitles)], 3)
   } else {
     coltitles <- c(paste0("Folder", paste(rep(" ", len - 6), collapse = "")),
-      "Runtime    ", "RunType    ", "RunStatus        ", "Mif",
+      "Runtime    ", "RunType    ", "RunStatus        ", "Mif   ",
       "Conv                 ", "Iter            ", "modelstat          ")
     lenCols <- nchar(coltitles)
   }

R/modeltests.R

@@ -267,7 +267,7 @@ evaluateRuns <- function(model, mydir, gitPath, compScen, email, mattermostToken
   colSep <- "  "
   coltitles <- c(
     "Run                                           ", "Runtime    ", "", "RunType    ", "RunStatus         ",
-    "Iter            ", "Conv                 ", "modelstat          ", "Mif", "AppResults"
+    "Iter            ", "Conv                 ", "modelstat          ", "Mif   ", "AppResults"
   )
   write(paste(coltitles, collapse = colSep), myfile, append = TRUE)
   lenCols <- c(nchar(coltitles)[-length(coltitles)], 3)
@@ -289,6 +289,9 @@ evaluateRuns <- function(model, mydir, gitPath, compScen, email, mattermostToken
       if (grepl("testOneRegi", grsi[, "RunType"]) && grsi[, "modelstat"] != "2: Locally Optimal") {
         errorList <- c(errorList, "testOneRegi does not return an optimal solution")
       }
+      if (grsi[, "Mif"] == "sumErr") {
+        errorList <- c(errorList, "Summation checks for some run(s) revealed some gaps")
+      }
     } else if (model == "MAgPIE") {
       if (paste0(unique(unlist(strsplit(gsub("[^0-9]", "", grsi[, "modelstat"]), split = ""))), collapse = "") != "2") {
         errorList <- c(errorList, "Some run(s) did not converge")
@@ -305,15 +308,15 @@ evaluateRuns <- function(model, mydir, gitPath, compScen, email, mattermostToken
       # Doesn't this check exclude runs that have 'converged (had INFES)' and 
       # wouldn't it be useful to also produce compareScenarios for these? The existence of the mif file only
       # matters later on after the runtime check when it comes to compareScenarios. So why testing for it here?
-      if (any(grepl(basename(getwd()), rownames(filter(gRS, .data$Conv == "converged", .data$Mif == "yes"))))) {
+      if (any(grepl(basename(getwd()), rownames(filter(gRS, .data$Conv == "converged", .data$Mif %in% c("yes", "sumErr")))))) {
         load("config.Rdata")
       } else {
         setwd("../")
         message("Skipping ", i, " and changed back to ", normalizePath("."))
         next
       }
       sameRuns <- gRS %>% filter(.data$Conv %in% c("converged", "converged (had INFES)"), # runs have to be converged
-                                 .data$Mif == "yes",                                      # need to have mifs
+                                 .data$Mif %in% c("yes", "sumErr"),                       # need to have mifs
                                  grepl(cfg$title, rownames(gRS)),                         # must be the same scenario
                                  ! rownames(gRS) %in% basename(cfg$results_folder)) %>%   # but not the current run
                           rownames()
@@ -403,7 +406,7 @@ evaluateRuns <- function(model, mydir, gitPath, compScen, email, mattermostToken
         }
       }
     }
-    write(paste0("The IAMC check of these runs can be found in /p/projects/remind/modeltests/output/iamccheck-",
+    write(paste0("The IAMC check of these runs can be found in /p/projects/remind/modeltests/remind/output/iamccheck-",
                  commitTested, ".rds", "\n"), myfile, append = TRUE)
   }
 

README.md

@@ -1,6 +1,6 @@
 # Run Analysis Tools
 
-R package **modelstats**, version **0.18.1**
+R package **modelstats**, version **0.19.0**
 
 [![CRAN status](https://www.r-pkg.org/badges/version/modelstats)](https://cran.r-project.org/package=modelstats)  [![R build status](https://github.com/pik-piam/modelstats/workflows/check/badge.svg)](https://github.com/pik-piam/modelstats/actions) [![codecov](https://codecov.io/gh/pik-piam/modelstats/branch/master/graph/badge.svg)](https://app.codecov.io/gh/pik-piam/modelstats) [![r-universe](https://pik-piam.r-universe.dev/badges/modelstats)](https://pik-piam.r-universe.dev/builds)
 
@@ -47,7 +47,7 @@ In case of questions / problems please contact Anastasis Giannousakis <giannou@p
 
 To cite package **modelstats** in publications use:
 
-Giannousakis A, Richters O (2024). _modelstats: Run Analysis Tools_. R package version 0.18.1, <https://github.com/pik-piam/modelstats>.
+Giannousakis A, Richters O (2024). _modelstats: Run Analysis Tools_. R package version 0.19.0, <https://github.com/pik-piam/modelstats>.
 
 A BibTeX entry for LaTeX users is
 
@@ -56,7 +56,7 @@ A BibTeX entry for LaTeX users is
   title = {modelstats: Run Analysis Tools},
   author = {Anastasis Giannousakis and Oliver Richters},
   year = {2024},
-  note = {R package version 0.18.1},
+  note = {R package version 0.19.0},
   url = {https://github.com/pik-piam/modelstats},
 }
 ```