From da54a4064c52c13a584af34dd194d3c214d87319 Mon Sep 17 00:00:00 2001
From: Al-Murphy <alanmurph94@hotmail.com>
Date: Fri, 28 Jul 2023 10:46:01 +0200
Subject: [PATCH] bug fix: imputation ind with missing rs ids

---
 DESCRIPTION         |  2 +-
 NEWS.md             |  8 +++++
 R/check_no_rs_snp.R |  3 +-
 R/compute_nsize.R   | 84 +++++++++++++++++++++++----------------------
 4 files changed, 54 insertions(+), 43 deletions(-)

diff --git a/DESCRIPTION b/DESCRIPTION
index 95f3d64..6f5c71d 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,7 +1,7 @@
 Package: MungeSumstats
 Type: Package
 Title: Standardise summary statistics from GWAS
-Version: 1.9.14
+Version: 1.9.15
 Authors@R:
     c(person(given = "Alan",
            family = "Murphy",
diff --git a/NEWS.md b/NEWS.md
index 6a96724..ad57350 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,3 +1,11 @@
+## CHANGES IN VERSION 1.9.15
+
+### Bug fix
+* Fix for imputation column when imputing RS ID from CHR:BP. Avoids crash and 
+ensures correct identification of imputed SNPs.
+* Avoid running compute_nsize function when no imputation is wanted by user - 
+also avoids message output in this situation.
+
 ## CHANGES IN VERSION 1.9.14
 
 ### Bug fix
diff --git a/R/check_no_rs_snp.R b/R/check_no_rs_snp.R
index b0fa88c..71809f5 100644
--- a/R/check_no_rs_snp.R
+++ b/R/check_no_rs_snp.R
@@ -266,7 +266,8 @@ check_no_rs_snp <- function(sumstats_dt, path, ref_genome, snp_ids_are_rs_ids,
                 miss_rs_chr_bp[, (format) := NULL]
                 # join with full dataset
                 # If IMPUTATION column added add it to other DT
-                if (imputation_ind &&
+                if (imputation_ind && 
+                    "IMPUTATION_SNP" %in% names(sumstats_dt) &&
                     !"IMPUTATION_SNP" %in% names(miss_rs_chr_bp)) {
                     miss_rs_chr_bp[, IMPUTATION_SNP := NA]
                 }
diff --git a/R/compute_nsize.R b/R/compute_nsize.R
index 87085f8..a85a018 100644
--- a/R/compute_nsize.R
+++ b/R/compute_nsize.R
@@ -46,47 +46,49 @@ compute_nsize <- function(sumstats_dt,
                           return_list=TRUE) {
     ## Avoid confusing BiocCheck.
     IMPUTATION_n <- IMPUTATION_Neff <- NULL;
-    #### Copy data.table ####
-    ## IMPORTANT!: This makes it so that any subsequent changes made to 
-    ## the internal sumstats_dt variable (and the function output) will not 
-    ## alter the original input data as well. 
-    sumstats_dt <- data.table::copy(sumstats_dt) 
-    #### Standardise the column names before continuing ####
-    if (standardise_headers) {
-        sumstats_dt <-
-            standardise_sumstats_column_headers_crossplatform(
-                sumstats_dt = sumstats_dt, 
-            )[["sumstats_dt"]]
-    }
-    # if you want an Neff column for more than one method need to ensure
-    # you can tell which is which
-    append_method_name <- FALSE
-    if (is.vector(compute_n)) {
-        # sum makes an N column not an Neff column
-        if (length(compute_n[!compute_n == "sum"]) > 1) {
-              append_method_name <- TRUE
-          }
-    }
-    for (method in compute_n) {
-        compute_sample_size(
-            sumstats_dt = sumstats_dt,
-            method = method,
-            force_new = force_new,
-            append_method_name = append_method_name
-        )
-    }
-    # if user wants information, give SNPs where Z-score calculated
-    ## Evaluate the conditions under which N would have been calculate by
-    ## compute_sample_size()
-    if (imputation_ind &&
-        is.numeric(compute_n) && compute_n != 0L &&
-        (!"N" %in% names(sumstats_dt))) {
-        sumstats_dt[, IMPUTATION_n := TRUE]
-    }
-    if (imputation_ind &&
-        is.character(compute_n)) {
-        sumstats_dt[, IMPUTATION_Neff := TRUE]
-    }
+    if (!is.numeric(compute_n) || (is.numeric(compute_n) && compute_n != 0L)){
+      #### Copy data.table ####
+      ## IMPORTANT!: This makes it so that any subsequent changes made to 
+      ## the internal sumstats_dt variable (and the function output) will not 
+      ## alter the original input data as well. 
+      sumstats_dt <- data.table::copy(sumstats_dt) 
+      #### Standardise the column names before continuing ####
+      if (standardise_headers) {
+          sumstats_dt <-
+              standardise_sumstats_column_headers_crossplatform(
+                  sumstats_dt = sumstats_dt, 
+              )[["sumstats_dt"]]
+      }
+      # if you want an Neff column for more than one method need to ensure
+      # you can tell which is which
+      append_method_name <- FALSE
+      if (is.vector(compute_n)) {
+          # sum makes an N column not an Neff column
+          if (length(compute_n[!compute_n == "sum"]) > 1) {
+                append_method_name <- TRUE
+            }
+      }
+      for (method in compute_n) {
+          compute_sample_size(
+              sumstats_dt = sumstats_dt,
+              method = method,
+              force_new = force_new,
+              append_method_name = append_method_name
+          )
+      }
+      # if user wants information, give SNPs where Z-score calculated
+      ## Evaluate the conditions under which N would have been calculate by
+      ## compute_sample_size()
+      if (imputation_ind &&
+          is.numeric(compute_n) && compute_n != 0L &&
+          (!"N" %in% names(sumstats_dt))) {
+          sumstats_dt[, IMPUTATION_n := TRUE]
+      }
+      if (imputation_ind &&
+          is.character(compute_n)) {
+          sumstats_dt[, IMPUTATION_Neff := TRUE]
+      }
+    }  
     #### Return format ####
     if(return_list){
         return(list("sumstats_dt" = sumstats_dt))