Optimizations for IAF (#553) (#559)

Co-authored-by: Adam Wenocur <[email protected]> Co-authored-by: Adam Wenocur <[email protected]>
TileDB-Inc · Aug 7, 2023 · 16387a3 · 16387a3
1 parent c545215
commit 16387a3
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 6 deletions.
diff --git a/apis/python/tests/test_tiledbvcf.py b/apis/python/tests/test_tiledbvcf.py
@@ -1177,7 +1177,7 @@ def test_ingest_with_stats(tmp_path):
         data_frame[
             (data_frame["sample_name"] == "second") & (data_frame["pos_start"] == 4)
         ]["info_TILEDB_IAF"].iloc[0][0]
-        == 0.0625
+        == 0.125
     )
 
 

diff --git a/libtiledbvcf/src/read/reader.cc b/libtiledbvcf/src/read/reader.cc
@@ -1246,6 +1246,10 @@ bool Reader::process_query_results_v4() {
   const auto& regions = regions_indexes->second;
 
   bool apply_af_filter = af_filter_enabled();
+  size_t num_samples = 0;
+  if (params_.scan_all_samples) {
+    num_samples = dataset_->sample_names().size();
+  }
   for (; read_state_.cell_idx < num_cells; read_state_.cell_idx++) {
     // For easy reference
     const uint64_t i = params_.sort_real_start_pos ?
@@ -1283,10 +1287,6 @@ bool Reader::process_query_results_v4() {
 
       read_state_.query_results.af_values.clear();
       int allele_index = 0;
-      size_t num_samples = 0;
-      if (params_.scan_all_samples) {
-        num_samples = dataset_->sample_names().size();
-      }
       bool is_ref = true;
       for (auto&& allele : alleles) {
         auto [allele_passes, af] = af_filter_->pass(

diff --git a/libtiledbvcf/src/stats/variant_stats_reader.h b/libtiledbvcf/src/stats/variant_stats_reader.h
@@ -137,7 +137,7 @@ class AFMap {
     if (next_allele == pos_map.second.end()) {
       return 0;
     }
-    return 1.0 * next_allele->second / pos_map.first;
+    return 1.0 * next_allele->second / num_samples;
   }
 
   /**