Merge pull request #1 from qclayssen/issue-20-vizualization-subsampling

Enhancements to Subsampling Feature in tidygate
stemangiola · Nov 27, 2023 · c93b37f · c93b37f
2 parents f05505d + 3106cc8
commit c93b37f
Show file tree

Hide file tree

Showing 9 changed files with 73 additions and 3 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -20,7 +20,7 @@ Depends:
     R (>= 3.6.0)
 Encoding: UTF-8
 LazyData: true
-RoxygenNote: 7.1.1
+RoxygenNote: 7.2.3
 Imports: 
     utils,
     graphics,

diff --git a/R/functions_chr_int.R b/R/functions_chr_int.R
@@ -45,6 +45,7 @@ pretty_plot_chr_int = function(.data,
                                .color = NULL,
                                .shape = NULL,
                                .size = NULL,
+                               subsample = NULL,
                                opacity = 1,
                                is_size_fixed) {
   # Comply with CRAN NOTES
@@ -58,8 +59,19 @@ pretty_plot_chr_int = function(.data,
   .shape = enquo(.shape)
   .size = enquo(.size)
   my_size_range = c(1, 3)
-
-
+
+  if (!is.null(subsample)) {
+      if (subsample > 0 && subsample < 1) {
+          # Subsample as a fraction
+          set.seed(123)  # Set seed for reproducibility
+          .data <- .data %>% sample_frac(subsample)
+      } else if (subsample >= 1) {
+          # Subsample as a fixed number
+          set.seed(123)  # Set seed for reproducibility
+          .data <- .data %>% sample_n(subsample)
+      }
+  }
+
 
   .data_formatted =
     .data %>%
@@ -263,6 +275,7 @@ pretty_plot_chr_int = function(.data,
 #' @param .color A column symbol. Color of points
 #' @param .shape A column symbol. Shape of points
 #' @param .size A column symbol. Size of points
+#' @param subsample A numeric value or a fraction indicating the subset of data to sample.
 #' @param opacity A number between 0 and 1. The opacity level of the data points
 #' @param how_many_gates An integer. The number of gates to label
 #' @param gate_list A list of gates. It is returned by gate function as attribute \"gate\". If you want to create this list yourself, each element of the list is a data frame with x and y columns. Each row is a coordinate. The order matter.
@@ -277,6 +290,7 @@ gate_interactive_chr_int <-
            .color = NA,
            .shape = NULL,
            .size = NULL,
+           subsample = NULL,
            opacity = 1,
            how_many_gates = 1,
            is_size_fixed,
@@ -336,6 +350,8 @@ gate_interactive_chr_int <-
       # size can be number or column
       .size = !!.size,
 
+      subsample = subsample,
+
       opacity = opacity,
       is_size_fixed = is_size_fixed
     )
@@ -427,6 +443,7 @@ gate_programmatic_chr_int <-
                            .color = NULL,
                            .shape = NULL,
                            .size = NULL,
+                           subsample = NULL,
                            opacity = 1,
                            how_many_gates = 1,
                            .group_by = NULL,
@@ -484,6 +501,7 @@ gate_programmatic_chr_int <-
       .color = .color,
       .shape = .shape,
 
+      subsample =subsample,
       # size can be number of column
       .size =  .size,
 

diff --git a/R/methods.R b/R/methods.R
@@ -16,6 +16,7 @@
 #' @param .color A column symbol. Colour of points
 #' @param .shape A column symbol. Shape of points
 #' @param .size A column symbol. Size of points
+#' @param subsample A numeric value or a fraction indicating the subset of data to sample.
 #' @param opacity A number between 0 and 1. The opacity level of the data points
 #' @param how_many_gates An integer. The number of gates to label
 #' @param .group_by A column symbol. The column that is used to calculate distance (i.e., normally genes)
@@ -73,6 +74,7 @@ gate_chr <- function(.dim1,
                      .color = NULL,
                      .shape = NULL,
                      .size = NULL,
+                     subsample = NULL,
                      opacity = 1,
                      how_many_gates = 1,
                      .group_by = NULL,
@@ -95,6 +97,7 @@ gate_chr.numeric = 	function(                     .dim1,
                                                    .color = NULL,
                                                    .shape = NULL,
                                                    .size = NULL,
+                                                   subsample = NULL,
                                                    opacity = 1,
                                                    how_many_gates = 1,
                                                    .group_by = NULL,
@@ -109,6 +112,7 @@ gate_chr.numeric = 	function(                     .dim1,
     .color = .color,
     .shape = .shape,
     .size = .size,
+    subsample = subsample,
     opacity = opacity,
     how_many_gates = how_many_gates,
     .group_by = .group_by,
@@ -136,6 +140,7 @@ gate_int <- function(.dim1,
                      .color = NULL,
                      .shape = NULL,
                      .size = NULL,
+                     subsample = NULL,
                      opacity = 1,
                      how_many_gates = 1,
                      .group_by = NULL,
@@ -157,6 +162,7 @@ gate_int.numeric = 	function(  .dim1,
                                                    .color = NULL,
                                                    .shape = NULL,
                                                    .size = NULL,
+                                                   subsample = NULL,
                                                    opacity = 1,
                                                    how_many_gates = 1,
                                                    .group_by = NULL,
@@ -170,6 +176,7 @@ gate_int.numeric = 	function(  .dim1,
     .color = .color,
     .shape = .shape,
     .size = .size,
+    subsample = subsample,
     opacity = opacity,
     how_many_gates = how_many_gates,
     .group_by = .group_by,

diff --git a/README.Rmd b/README.Rmd
@@ -118,4 +118,15 @@ tidygate_data %>%
 
 ```
 
+## Subsampling
 
+The `subsample` parameter allows you to sample a subset of your data for analysis with a fixed seed. This can be particularly useful for large datasets or for preliminary exploratory analysis. Here's how you can use it:
+
+```{r}
+# For sampling 50% of your data
+result <- your_function(data, subsample = 0.5)
+
+# For sampling 100 specific observations
+result <- your_function(data, subsample = 100)
+
+```
diff --git a/README.md b/README.md
@@ -131,3 +131,24 @@ tidygate_data %>%
     ##  9 adrenal         1 fibrobla… immune_c…     0.52 ACC       -0.776 -0.383  0    
     ## 10 adrenal         1 immune_c… endothel…     1    ACC        0.980 -0.116  0    
     ## # … with 2,230 more rows
+
+## Subsampling
+
+The `subsample` parameter allows you to sample a subset of your data for analysis with a fixed seed. This can be particularly useful for large datasets or for preliminary exploratory analysis. Here's how you can use it:
+
+```r
+# For sampling 50% of your data
+tidygate_data %>%
+  mutate( gate = gate_chr(
+    Dim1, Dim2, subsample = 0.5
+  ))
+
+
+# For sampling 100 specific observations
+tidygate_data %>%
+  mutate( gate = gate_chr(
+    Dim1, Dim2, subsample = 100
+  ))
+
+
+```
diff --git a/man/gate_chr-methods.Rd b/man/gate_chr-methods.Rd
diff --git a/man/gate_chr.numeric.Rd b/man/gate_chr.numeric.Rd
diff --git a/man/gate_int.numeric.Rd b/man/gate_int.numeric.Rd
diff --git a/man/gate_interactive_chr_int.Rd b/man/gate_interactive_chr_int.Rd