Merge pull request #131 from mountainMath/0.4.1

fix problem with parsing census tables, add unit test

Author: mountainMath

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: cansim
 Type: Package
 Title: Accessing Statistics Canada Data Table and Vectors
-Version: 0.4
+Version: 0.4.1
 Authors@R: c(
     person("Jens", "von Bergmann", email = "jens@mountainmath.ca", role = c("aut","cre")),
     person("Dmitry", "Shkolnik", email = "shkolnikd@gmail.com", role = c("aut")))

NEWS.md

@@ -1,3 +1,8 @@
+# cansim 0.4.1
+## Minor changes
+* fix problem with parsing census data tables
+* fix problem with converting to factors when classification codes are attached.
+
 # cansim 0.4
 ## Major changes
 * add support for local caching in parquet and feather formats

R/cansim.R

@@ -8,7 +8,8 @@
 #' @param default_month The default month that should be used when creating Date objects for annual data (default set to "01")
 #' @param default_day The default day of the month that should be used when creating Date objects for monthly data (default set to "01")
 #' @param factors (Optional) Logical value indicating if dimensions should be converted to factors. (Default set to \code{TRUE}).
-#' @param strip_classification_code (strip_classification_code) Logical value indicating if classification code should be stripped from names. (Default set to \code{false}).
+#' @param strip_classification_code (strip_classification_code) Logical value indicating if classification code should be stripped
+#' from names. (Default set to \code{FALSE}, if \code{factors=TRUE} this is overridden and set to \code{TRUE}).
 #' @param cansimTableNumber (Optional) Only needed when operating on results of SQLite connections.
 #' @param internal (Optional) Flag to indicate that this function is called internally.
 #'
@@ -136,6 +137,9 @@ normalize_cansim_values <- function(data, replacement_value="val_norm", normaliz
     fields= gsub(classification_prefix,"",names(data)[grepl(classification_prefix,names(data))])
   }
 
+  if (factors) { #override for consistency
+    strip_classification_code=TRUE
+  }
 
   if (strip_classification_code){
     for (field in fields) {
@@ -203,7 +207,7 @@ normalize_cansim_values <- function(data, replacement_value="val_norm", normaliz
   }
 
   # column order
-  if (replacement_value_string != value_string) {
+  if (replacement_value_string != value_string && replacement_value_string %in% names(data) && value_string %in% names(data)) {
     data <- data %>%
       relocate(!!as.name(replacement_value_string),.after=!!as.name(value_string))
   }
@@ -489,6 +493,9 @@ get_cansim <- function(cansimTableNumber, language="english", refresh=FALSE, tim
     data <- readRDS(file=data_path)
   }
 
+  attr(data,"language") <- cleaned_language
+  attr(data,"cansimTableNumber") <- cleaned_number
+
   if (!is.null(getOption("cansim.debug"))) message('Initiating normalization')
   data <- data  %>%
     normalize_cansim_values(replacement_value = "val_norm", factors = factors,

R/cansim_members.R

@@ -28,7 +28,7 @@ get_cansim_members <- function(cansimTableNumber, language="english",
     })
 
     members %>%
-    seq_along() %>%
+    #seq_along() %>%
     lapply(\(i,el,n){
 
     })

R/cansim_parquet.R

@@ -518,7 +518,7 @@ collect_and_normalize <- function(connection,
   cansimTableNumber <- attr(connection,"cansimTableNumber")
   language <- attr(connection,"language")
 
-  if ((is.null(language)||is.null(cansimTableNumber))&& ("arrow_dplyr_query" %in% class(connection))) {
+  if ((is.null(language)||is.null(cansimTableNumber))&& ("arrow_dplyr_query" %in% class(connection) || "ArrowObject" %in% class(connection))) {
     md <- arrow::schema(connection)$metadata
     cansimTableNumber <- md$cansimTableNumber
     language <- md$language
@@ -536,7 +536,7 @@ collect_and_normalize <- function(connection,
     attr(data,"cansimTableNumber") <- cansimTableNumber
 
     if (disconnect) disconnect_cansim_sqlite(connection)
-  } else if ("arrow_dplyr_query" %in% class(connection)){
+  } else if ("arrow_dplyr_query" %in% class(connection) || "ArrowObject" %in% class(connection)){
     data <- connection %>%
       dplyr::as_tibble()
     attr(data,"language") <- language

README.md

@@ -231,7 +231,7 @@ If you want to get in touch, we are pretty good at responding via email or via t
 
 If you wish to cite the `cansim` package in your work:
 
-  von Bergmann, J., Dmitry Shkolnik (2024). cansim: functions and convenience tools for accessing Statistics Canada data tables. v0.4. DOI: 10.32614/CRAN.package.cansim
+  von Bergmann, J., Dmitry Shkolnik (2024). cansim: functions and convenience tools for accessing Statistics Canada data tables. v0.4.1. DOI: 10.32614/CRAN.package.cansim
 
 A BibTeX entry for LaTeX users is
 
@@ -241,7 +241,7 @@ A BibTeX entry for LaTeX users is
     title = {cansim: functions and convenience tools for accessing Statistics Canada data tables},
     year = {2025},
     doi = {10.32614/CRAN.package.cansim},
-    note = {R package version 0.4},
+    note = {R package version 0.4.1},
     url = {https://mountainmath.github.io/cansim/}
   }
 ```

cran-comments.md

@@ -125,3 +125,9 @@ There were no ERRORs or WARNINGs or NOTEs.
 * fix problem with inconsistent type parsing of notes
 * better support for french language when accessing data by vector or coordinate
 * tests
+
+# cansim 0.4.1
+## Minor changes
+* fix problem with parsing census data tables
+* fix problem with converting to factors when classification codes are attached.
+