From 9dd94755b9e1213b80825b00aeb15e3777305409 Mon Sep 17 00:00:00 2001
From: Ally Hawkins <ally.hawkins@ccdatalab.org>
Date: Thu, 12 Dec 2024 14:46:37 -0600
Subject: [PATCH 1/2] fix join error with cluster column

---
 analyses/cell-type-ewings/scripts/utils/clustering-functions.R | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/analyses/cell-type-ewings/scripts/utils/clustering-functions.R b/analyses/cell-type-ewings/scripts/utils/clustering-functions.R
index 5c0f01bfb..3df5dd415 100644
--- a/analyses/cell-type-ewings/scripts/utils/clustering-functions.R
+++ b/analyses/cell-type-ewings/scripts/utils/clustering-functions.R
@@ -56,7 +56,7 @@ get_cluster_stats <- function(sce,
       return(stats_df)
     }) |>
     dplyr::bind_rows(.id = "cluster_params") |> 
-    dplyr::left_join(all_cluster_results, by = c("cell_id", "cluster_params"))
+    dplyr::left_join(all_cluster_results, by = c("cell_id", "cluster_params", "cluster"))
 
   return(all_stats_df)
 }

From 5e7f4ccd82277444cc1aaa56725a61115b3879f2 Mon Sep 17 00:00:00 2001
From: Ally Hawkins <ally.hawkins@ccdatalab.org>
Date: Thu, 12 Dec 2024 14:46:47 -0600
Subject: [PATCH 2/2] add table with number of clusters

---
 .../01-clustering-metrics.Rmd                 | 27 ++++++++++++++-----
 1 file changed, 21 insertions(+), 6 deletions(-)

diff --git a/analyses/cell-type-ewings/template_notebooks/clustering-workflow/01-clustering-metrics.Rmd b/analyses/cell-type-ewings/template_notebooks/clustering-workflow/01-clustering-metrics.Rmd
index 49be4bf2c..84d4c1447 100644
--- a/analyses/cell-type-ewings/template_notebooks/clustering-workflow/01-clustering-metrics.Rmd
+++ b/analyses/cell-type-ewings/template_notebooks/clustering-workflow/01-clustering-metrics.Rmd
@@ -149,6 +149,21 @@ all_stats <- split_umap_df |>
   purrr::map(\(df) {get_cluster_stats(sce,df)})
 ```
 
+Before we show any statistics, we will first show the number of clusters for each set of parameters. 
+For any cases where we only see 1 cluster, no distributions will be shown in the plots. 
+
+```{r}
+all_stats |> 
+  purrr::imap(\(df, method){
+    df |> 
+      dplyr::select(resolution, nn, cluster) |> 
+      dplyr::group_by(resolution, nn) |> 
+      dplyr::summarise(num_of_clusters = length(unique(cluster))) |>
+      knitr::kable()
+  })
+```
+
+
 ### Silhouette width
 
 ```{r}
@@ -169,17 +184,17 @@ all_stats |>
 
 ```{r, warning=FALSE, message=FALSE}
 # calculate cluster stability
-stability_stats <- split_umap_df |> 
+stability_stats <- split_umap_df |>
   purrr::map(\(df){ get_cluster_stability(sce, df, params$threads)})
 ```
 
 
 ```{r, warning=FALSE, message=FALSE}
 # plot stability across all parameters
-stability_stats |> 
+stability_stats |>
   purrr::imap(\(df, cluster_method){
-    
-    
+
+
     # plot stability across all values of k
     ggplot(df, aes(x = nn, y = ari)) +
       geom_jitter(width = 0.1) +
@@ -198,8 +213,8 @@ stability_stats |>
           quantile(x, 0.75)
         }
       )
-    
-    
+
+
   })
 ```