Merge branch 'feat@parallel' of github.com:tanaylab/tglkmeans into fe…

…at@parallel

NEWS.md

@@ -8,15 +8,15 @@
 
 * Added validity checks for `k` and the number of observations. 
 
-# tgkmeans 0.3.11
+# tglkmeans 0.3.11
 
 * Changed pkgdoc, see: https://github.com/r-lib/roxygen2/issues/1491.
 
-# tgkmeans 0.3.10
+# tglkmeans 0.3.10
 
 * Removed broken link to one of the references in the description.
 
-# tgkmeans 0.3.9
+# tglkmeans 0.3.9
 
 * Remove empty clusters. This may happen when the number of clusters is larger than the number of observations, and currently caused an error in the reordering step.
 

R/TGL_kmeans.R

@@ -99,6 +99,15 @@ TGL_kmeans_tidy <- function(df,
 
     mat <- df
 
+    # make sure that the input is numeric
+    mat <- as.matrix(mat)
+    if (!is.numeric(mat)) {
+        if (any(!is.numeric(mat[, 1])) && missing(id_column) && !id_column) {
+            cli_abort("{.field df} must be numeric. Note that the default of {.field id_column} was changed to FALSE in version {.field 0.4.0}. If you want to use the first column as ids, please set {.field id_column=TRUE}")
+        }
+        cli_abort("{.field df} must be numeric.")
+    }
+
     if (k < 1) {
         cli_abort("k must be greater than 0")
     }

README.Rmd

@@ -81,4 +81,4 @@ browseVignettes("usage")
 
 ## A note regarding random number generation
 
-From version 0.4.0 onwards, the package uses R random number generation functions instead of the C++11 random number generation functions. Note that this may result in different results from previous versions. To get the same results as previous versions, set the `use_cpp_random` argument to `TRUE` in the `TGL_kmeans` function.
+From version 0.4.0 onward, the package uses R random number generation functions instead of the C++11 random number generation functions. Note that this may result in different results from previous versions. To get the same results as previous versions, set the `use_cpp_random` argument to `TRUE` in the `TGL_kmeans` function.

README.md

@@ -123,7 +123,7 @@ browseVignettes("usage")
 
 ## A note regarding random number generation
 
-From version 0.4.0 onwards, the package uses R random number generation
+From version 0.4.0 onward, the package uses R random number generation
 functions instead of the C++11 random number generation functions. Note
 that this may result in different results from previous versions. To get
 the same results as previous versions, set the `use_cpp_random` argument

src/AParamStat.h

@@ -55,7 +55,7 @@ float wilcoxon_rank_sum(list <T> &samples, int type = 1) {
     float EU = n1 * n2 / 2.0;
     float VarU = n1 * n2 * (samples.size() + 1) / 12.0;
 
-    Rcpp::Rcout << "W " << W << " n2 " << n2 << " EU " << EU << " Var " << VarU << endl;
+    Rcpp::Rcout << "W " << W << " n2 " << n2 << " EU " << EU << " Var " << VarU << " t2_minus_t " << t3_minus_t << endl;
 
     float pv = erfc((U - EU) / sqrt(VarU));
 
@@ -124,7 +124,7 @@ float siegel_tukey(list <T> &samples, int type = 1) {
     float EU = n1 * n2 / 2.0;
     float VarU = n1 * n2 * (samples.size() + 1) / 12.0;
 
-    Rcpp::Rcout << "W " << W << " n2 " << n2 << " EU " << EU << " Var " << VarU << endl;
+    Rcpp::Rcout << "W " << W << " n2 " << n2 << " EU " << EU << " Var " << VarU << " t2_minus_t " << t3_minus_t << endl;
     float pv = erfc((U - EU) / sqrt(VarU));
 
     return (pv);

vignettes/usage.Rmd

@@ -19,6 +19,7 @@ library(dplyr)
 library(ggplot2)
 library(tglkmeans)
 theme_set(theme_classic())
+set.seed(60427)
 ```
 
 First, let's create 5 clusters normally distributed around 1 to 5, with sd of 0.3: