merge

Merge branch 'master' into github-pages-for-cmdstanr

# Conflicts:
#	DESCRIPTION

Author: stemangiola

Diff for: .Rbuildignore

@@ -18,11 +18,10 @@ src/Rcpp*
 ^README*
 ^dev
 ^rsconnect
+# DO NOT ADD ANY INST FILES HERE AS THE INST DIRECTORY WILL NOT BE INSTALLED
 model_glm_dirichlet_multinomial.rds
 model_glm_dirichlet_multinomial_generate_quantities.rds
 model_glm_dirichlet_multinomial_imputation.rds
-inst/stan/glm_multi_beta_binomial
-inst/stan/glm_multi_beta_binomial_generate_data
 ^doc$
 ^Meta$
 sccomp_draws_files

Diff for: DESCRIPTION

@@ -1,6 +1,6 @@
 Package: sccomp
 Title: Tests differences in cell-type proportion for single-cell data, robust to outliers
-Version: 1.99.9
+Version: 1.99.14
 Authors@R: c(person("Stefano", "Mangiola", email = "[email protected]",
                   role = c("aut", "cre"))
                   )
@@ -34,7 +34,8 @@ Imports:
     readr,
     scales,
     stringr,
-    glue
+    glue,
+    crayon
 Suggests:
     knitr,
     rmarkdown,

Diff for: NAMESPACE

@@ -34,9 +34,12 @@ import(dplyr)
 import(ggplot2)
 importFrom(SingleCellExperiment,colData)
 importFrom(SingleCellExperiment,counts)
+importFrom(crayon,red)
+importFrom(dplyr,add_count)
 importFrom(dplyr,any_of)
 importFrom(dplyr,arrange)
 importFrom(dplyr,as_tibble)
+importFrom(dplyr,count)
 importFrom(dplyr,cummean)
 importFrom(dplyr,distinct)
 importFrom(dplyr,distinct_at)
@@ -97,6 +100,7 @@ importFrom(readr,read_file)
 importFrom(readr,type_convert)
 importFrom(readr,write_file)
 importFrom(rlang,":=")
+importFrom(rlang,abort)
 importFrom(rlang,check_installed)
 importFrom(rlang,enquo)
 importFrom(rlang,ensym)
@@ -140,4 +144,3 @@ importFrom(tidyselect,all_of)
 importFrom(tidyselect,eval_select)
 importFrom(tidyselect,last_col)
 importFrom(utils,data)
-importFrom(utils,menu)

Diff for: R/functions_multi_beta_binomial.R

@@ -139,14 +139,16 @@ sccomp_glm_data_frame_counts = function(.data,
   .sample_cell_group_pairs_to_exclude = enquo(.sample_cell_group_pairs_to_exclude)
   #.grouping_for_random_effect = enquo(.grouping_for_random_effect)
 
+  # Check Sample Consistency of Factors
+  check_sample_consistency_of_factors(.data, formula_composition, !!.sample, !!.cell_group)
 
   #Check column class
   check_if_columns_right_class(.data, !!.sample, !!.cell_group)
 
   # Check that count is integer
   if(.data %>% pull(!!.count) %>% is("integer")) message(sprintf("sccomp says: %s column is an integer. The sum-constrained beta binomial model will be used", quo_name(.count)))
   else if(.data %>% pull(!!.count) %>% is("integer") |> not() & .data %>% pull(!!.count) |> dplyr::between(0, 1) |> all()) message(sprintf("sccomp says: %s column is a proportion. The sum-constrained beta model will be used. When possible using counts is preferred as the binomial noise component is often dominating for rare groups (e.g. rare cell types).", quo_name(.count)))
-  else stop(sprintf("sccomp: %s column must be an integer or a proportion", quo_name(.count)))
+  else stop(sprintf("sccomp: `%s` column must be an integer or a proportion", quo_name(.count)))
 
   # Check if columns exist
   check_columns_exist(.data, c(
@@ -240,7 +242,13 @@ sccomp_glm_data_frame_counts = function(.data,
 
   data_for_model =
     .data %>%
-    data_to_spread ( formula_composition, !!.sample, !!.cell_group, !!.count, .grouping_for_random_effect) %>%
+    data_to_spread ( formula_composition, !!.sample, !!.cell_group, !!.count, .grouping_for_random_effect) |>
+    
+    # This emerged with
+    # https://github.com/MangiolaLaboratory/sccomp/issues/175#issuecomment-2622749180
+    check_if_sample_is_a_unique_identifier(!!.sample) |> 
+    
+    # Create input for Stan
     data_spread_to_model_input(
       formula_composition, !!.sample, !!.cell_group, !!.count,
       truncation_ajustment = 1.1,

Diff for: R/methods.R

@@ -463,7 +463,6 @@ sccomp_estimate.data.frame <- function(.data,
     .abundance <- .count
   }
 
-  
   # DEPRECATION OF approximate_posterior_inference
   if (lifecycle::is_present(approximate_posterior_inference) & !is.null(approximate_posterior_inference)) {
     lifecycle::deprecate_warn("1.7.7", "sccomp::sccomp_estimate(approximate_posterior_inference = )", details = "The argument approximate_posterior_inference is now deprecated. Please use inference_method. By default, inference_method value is inferred from approximate_posterior_inference.")
@@ -1878,7 +1877,8 @@ sccomp_boxplot = function(
 
       pivot_wider(names_from = parameter, values_from = c(contains("c_"), contains("v_"))) |>
       unnest(count_data) |>
-      with_groups(!!.sample, ~ mutate(.x, proportion = (!!.count)/sum(!!.count)) ) 
+      with_groups(!!.sample, ~ mutate(.x, proportion = (!!.count)/sum(!!.count)) ) |> 
+      mutate(is_zero = proportion==0) 
 
   if(remove_unwanted_effects){
     .data_adjusted = 
@@ -1891,7 +1891,8 @@ sccomp_boxplot = function(
       select(-proportion) |> 
       left_join(.data_adjusted, by = join_by(!!.cell_group, !!.sample))
   }
-   
+  else 
+    message( "sccomp says: When visualising proportions, especially for complex models, consider setting `remove_unwanted_effects=TRUE`. This will adjust the proportions, preserving only the observed effect.")
 
   # If I don't have outliers add them
   if(!"outlier" %in% colnames(data_proportion)) data_proportion = data_proportion |> mutate(outlier = FALSE) 
@@ -2014,13 +2015,9 @@ else {
         # If discrete
         else 
           my_plot = 
-            plot_boxplot(
+            sccomp_boxplot(
               .data = x,
-              data_proportion = data_proportion,
-              factor_of_interest = .x,
-              .cell_group = !!.cell_group,
-              .sample =  !!.sample,
-              my_theme = multipanel_theme,
+              factor = .x,
               significance_threshold = significance_threshold
             )