* tidy_dtm/tidy_tdm did not order unnamed matrices as expected (e.g.,

  `{1, 2, ..., 1}` was ordered as `{1, 10, 2, ...}`).  This has been corrected.

close #16

Author: trinker

DESCRIPTION

@@ -1,10 +1,10 @@
 Package: textshape
-Date: 2018-01-27
+Date: 2018-02-11
 Title: Tools for Reshaping Text
 Version: 1.5.2
-Authors@R: c(person("Tyler", "Rinker", email = "tyler.rinker@gmail.com", role = c("aut", "cre")),
-            person("Joran", "Elias", role = "ctb"), person("Matthew", "Flickinger", role = "ctb"),
-            person('Paul', 'Foster', role = "ctb"))
+Authors@R: c(person("Tyler", "Rinker", email = "tyler.rinker@gmail.com", role = c("aut", "cre")), person("Joran",
+             "Elias", role = "ctb"), person("Matthew", "Flickinger", role = "ctb"), person('Paul', 'Foster', role =
+             "ctb"))
 Maintainer: Tyler Rinker <tyler.rinker@gmail.com>
 Description: Tools that can be used to reshape and restructure text data.
 Depends: R (>= 3.4.0)
@@ -16,3 +16,36 @@ Roxygen: list(wrap = FALSE)
 RoxygenNote: 6.0.1
 URL: http://github.com/trinker/textshape
 BugReports: http://github.com/trinker/textshape/issues
+Collate: 
+    'bind_list.R'
+    'bind_table.R'
+    'bind_vector.R'
+    'change_index.R'
+    'cluster_matrix.R'
+    'column_to_rownames.R'
+    'combine.R'
+    'duration.R'
+    'flatten.R'
+    'from_to.R'
+    'mtabulate.R'
+    'set_output.R'
+    'split_index.R'
+    'split_match.R'
+    'split_match_regex_to_transcript.R'
+    'split_portion.R'
+    'split_run.R'
+    'split_sentence.R'
+    'split_sentence_token.R'
+    'split_speaker.R'
+    'split_token.R'
+    'split_transcript.R'
+    'split_word.R'
+    'textshape-package.R'
+    'tidy_colo_dtm.R'
+    'utils.R'
+    'tidy_dtm.R'
+    'tidy_list.R'
+    'tidy_matrix.R'
+    'tidy_table.R'
+    'tidy_vector.R'
+    'unique_pairs.R'

NEWS

@@ -36,6 +36,9 @@ MINOR FEATURES
 
 IMPROVEMENTS
 
+* `tidy_dtm`/`tidy_tdm` did not order unnamed matrices as expected (e.g.,
+  `{1, 2, ..., 1}` was ordered as `{1, 10, 2, ...}`).  This has been corrected.
+
 CHANGES
 
 

NEWS.md

@@ -36,6 +36,9 @@ textshape 1.5.1 -
 
 **IMPROVEMENTS**
 
+* `tidy_dtm`/`tidy_tdm` did not order unnamed matrices as expected (e.g.,
+  `{1, 2, ..., 1}` was ordered as `{1, 10, 2, ...}`).  This has been corrected.
+
 **CHANGES**
 
 

R/tidy_dtm.R

@@ -10,6 +10,7 @@
 #' @param \ldots ignored.
 #' @return Returns a tidied data.frame.
 #' @rdname tidy_dtm
+#' @include utils.R
 #' @export
 #' @examples
 #' data(simple_dtm)
@@ -37,6 +38,7 @@
 #'     group_by(time) %>%
 #'     arrange(desc(n)) %>%
 #'     slice(1:10) %>%
+#'     ungroup() %>%
 #'     mutate(
 #'         term = factor(paste(term, time, sep = "__"),
 #'             levels = rev(paste(term, time, sep = "__")))
@@ -50,13 +52,15 @@ tidy_dtm <- function(x, ...){
 
     doc <- NULL
 
+    docfun <- function(docs) if (is_numeric_doc_names(x)) {as.integer(docs)} else {docs}
+
     data.table::data.table(
         doc = x[['dimnames']][['Docs']][x[['i']]],
         term = x[['dimnames']][['Terms']][x[['j']]],
         n = x[['v']],
         i = x[['i']],
         j = x[['j']]
-    )[order(doc), ]
+    )[, doc := docfun(doc)][order(doc), ][]
 }
 
 
@@ -67,11 +71,13 @@ tidy_tdm <- function(x, ...){
 
     doc <- NULL
 
+    docfun <- function(docs) if (is_numeric_doc_names(x)) {as.integer(docs)} else {docs}
+
     data.table::data.table(
         doc = x[['dimnames']][['Docs']][x[['j']]],
         term = x[['dimnames']][['Terms']][x[['i']]],
         n = x[['v']],
         i = x[['j']],
         j = x[['i']]
-    )[order(doc), ]
+    )[, doc := docfun(doc)][order(doc), ][]
 }

R/utils.R

@@ -117,3 +117,18 @@ sprintf_ish <- function(x, padding, type){
 #     set_tibble(x)
 # }
 
+is_numeric_doc_names <- function(x, ...){
+    UseMethod('is_numeric_doc_names')
+}
+
+
+is_numeric_doc_names.TermDocumentMatrix <- function(x, ...){
+    colnames_numeric <- suppressWarnings(as.integer(colnames(x)))
+    !anyNA(colnames_numeric) && isTRUE(all.equal(stats::sd(diff(colnames_numeric)), 0))
+}
+
+
+is_numeric_doc_names.DocumentTermMatrix <- function(x, ...){
+    rownames_numeric <- suppressWarnings(as.integer(rownames(x)))
+    !anyNA(rownames_numeric) && isTRUE(all.equal(stats::sd(diff(rownames_numeric)), 0))
+}